WO2020217494A1 - Fitting assistance device, fitting assistance method, and computer-readable recording medium - Google Patents

Fitting assistance device, fitting assistance method, and computer-readable recording medium Download PDF

Info

Publication number
WO2020217494A1
WO2020217494A1 PCT/JP2019/018077 JP2019018077W WO2020217494A1 WO 2020217494 A1 WO2020217494 A1 WO 2020217494A1 JP 2019018077 W JP2019018077 W JP 2019018077W WO 2020217494 A1 WO2020217494 A1 WO 2020217494A1
Authority
WO
WIPO (PCT)
Prior art keywords
information
parameter data
fitting
fitting support
output
Prior art date
Application number
PCT/JP2019/018077
Other languages
French (fr)
Japanese (ja)
Inventor
大 窪田
優太 芦田
英恵 下村
Original Assignee
日本電気株式会社
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 日本電気株式会社 filed Critical 日本電気株式会社
Priority to PCT/JP2019/018077 priority Critical patent/WO2020217494A1/en
Priority to JP2021515724A priority patent/JP7276433B2/en
Publication of WO2020217494A1 publication Critical patent/WO2020217494A1/en

Links

Images

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04R—LOUDSPEAKERS, MICROPHONES, GRAMOPHONE PICK-UPS OR LIKE ACOUSTIC ELECTROMECHANICAL TRANSDUCERS; DEAF-AID SETS; PUBLIC ADDRESS SYSTEMS
    • H04R25/00—Deaf-aid sets, i.e. electro-acoustic or electro-mechanical hearing aids; Electric tinnitus maskers providing an auditory perception
    • H04R25/70—Adaptation of deaf aid to hearing loss, e.g. initial electronic fitting
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04R—LOUDSPEAKERS, MICROPHONES, GRAMOPHONE PICK-UPS OR LIKE ACOUSTIC ELECTROMECHANICAL TRANSDUCERS; DEAF-AID SETS; PUBLIC ADDRESS SYSTEMS
    • H04R25/00—Deaf-aid sets, i.e. electro-acoustic or electro-mechanical hearing aids; Electric tinnitus maskers providing an auditory perception
    • H04R25/50—Customised settings for obtaining desired overall acoustical characteristics
    • H04R25/505—Customised settings for obtaining desired overall acoustical characteristics using digital signal processing
    • H04R25/507—Customised settings for obtaining desired overall acoustical characteristics using digital signal processing implemented by neural network or fuzzy logic

Definitions

  • the present invention relates to a fitting support device and a fitting support method, and further relates to a computer-readable recording medium that records a program for realizing these.
  • the mounting device used by mounting it on the living body needs to be adjusted (fitting) according to the living body.
  • a technician performs fitting in consideration of the subject's hearing, ear structure, living environment, and the like.
  • Patent Document 1 discloses a fitting device that automatically fits a hearing aid. According to the fitting device disclosed in Patent Document 1, the subject is made to audition the environmental sound created in advance, the subject is made to evaluate how to hear the auditioned environmental sound, and the hearing aid is used until the subject is satisfied. It is a device that fits by repeating the adjustment of.
  • Patent Document 1 can perform fitting so as not to feel the noise contained in the environmental sound noisy, it is not possible to carry out the fitting as performed by a highly skilled technician. difficult.
  • An example of an object of the present invention is to provide a fitting support device, a fitting support method, and a computer-readable recording medium for improving the accuracy of fitting.
  • the fitting support device in one aspect of the present invention is An acquisition means for acquiring inspection information representing the results of an inspection conducted on a subject and survey information representing the results of a survey related to the subject.
  • the acquired inspection information and the survey information are input to estimate the goodness of fit index representing the goodness of fit of the parameter data and its distribution.
  • Means and It is characterized by having.
  • the fitting support method in one aspect of the present invention is: (A) Obtain the inspection information indicating the result of the inspection conducted on the subject and the survey information representing the result of the investigation related to the subject. (B) In fitting the parameter data used to fit the device to the target person, the acquired inspection information and the survey information are input to estimate the goodness of fit index representing the goodness of fit of the parameter data and its distribution. It is characterized by doing.
  • a computer-readable recording medium on which a program according to one aspect of the present invention is recorded is On the computer (A) A step of acquiring inspection information representing the results of an inspection conducted on a subject and survey information representing the results of a survey related to the subject. (B) In fitting the parameter data used to fit the device to the target person, the acquired inspection information and the survey information are input to estimate the goodness of fit index representing the goodness of fit of the parameter data and its distribution. To do, step and It is characterized by recording a program including an instruction to execute.
  • the accuracy of fitting can be improved.
  • FIG. 1 is a diagram showing an example of a fitting support device.
  • FIG. 2 is a diagram showing an example of a system having a fitting support device.
  • FIG. 3 is a diagram showing an example of a data structure of selection information.
  • FIG. 4 is a diagram showing an example of a data structure of parameter data.
  • FIG. 5 is a diagram showing a display example of a conformity index of parameter data.
  • FIG. 6 is a diagram showing a display example of a conformity index of parameter data.
  • FIG. 7 is a diagram showing an example of a data structure of inspection information.
  • FIG. 8 is a diagram showing an example of the data structure of the survey information.
  • FIG. 9 is a diagram showing a display example of the amount of information corresponding to the missing information.
  • FIG. 10 is a diagram showing an example of a system having a learning device.
  • FIG. 11 is a diagram showing an example of the operation of the fitting support device.
  • FIG. 12 is a diagram showing an example of the operation of the learning device.
  • FIG. 13 is a diagram showing an example of a computer that realizes a fitting support device.
  • FIG. 1 is a diagram showing an example of a fitting support device.
  • the fitting support device shown in FIG. 1 is a device that improves the accuracy of adjustment (fitting) for adapting the device to the target person. Further, as shown in FIG. 1, the fitting support device 1 has an acquisition unit 2 and an estimation unit 3.
  • the device may be, for example, a device that is attached to a living body and used. Specifically, as devices, hearing aids, earphones, smart glasses, head-mounted displays, smart watches, musical instruments, medical devices, nursing robots, and the like can be considered. However, the device may be a device used without being attached to a living body.
  • the acquisition unit 2 acquires inspection information representing the results of the inspection conducted on the target person and survey information representing the results of the survey related to the target person.
  • the estimation unit 3 uses the acquired inspection information and the survey information as inputs to estimate the goodness of fit index representing the goodness of fit of the parameter data and its distribution.
  • the inspection information is information (explanatory variable) representing the result of the inspection conducted on the subject.
  • the survey information is information (explanatory variable) representing the results of the survey related to the subject.
  • Parameter data is data (set value) used to adjust the hardware and software of the target device.
  • the goodness of fit index is an index showing the goodness of fit of parameter data in fitting.
  • the conformity index for example, the probability of expressing the promising degree of the parameter data can be considered.
  • a number that meets constraints such as safety standards for fittings determined based on domain knowledge may be used as a conformity index.
  • the estimation unit 3 inputs a plurality of inspection information acquired in the past, a plurality of survey information acquired in the past, and parameter data set in the device in the past, and matches the parameter data generated by machine learning. It has a learning model used to estimate the index and its distribution.
  • Machine learning can be, for example, supervised learning or semi-supervised learning. Specifically, machine learning includes regression (regression method such as Bayesian linear regression), multiclass classification (algorithm such as decision tree), and the like.
  • regression regression method such as Bayesian linear regression
  • multiclass classification algorithm such as decision tree
  • the fitting support device 1 can estimate the matching index of the parameter data, so that the matching index of the estimated parameter data and its distribution can be presented to the technician. Therefore, the technician can carry out appropriate fitting for the target person.
  • the fitting support device 1 outputs a probability distribution to the output device as the distribution of the matching index of the parameter data, so that even a technician with a low skill level can refer to the output probability distribution and the target person. Appropriate fitting can be performed for. Therefore, the skill of a technician with a low skill level can be improved, and the accuracy of fitting can be improved.
  • FIG. 2 is a diagram showing an example of a system having a fitting support device 1.
  • the system having the fitting support device 1 in the present embodiment has an input device 21 and an output device 22 in addition to the fitting support device 1.
  • the system is not limited to the configuration shown in FIG.
  • the fitting support device 1 estimates (1) the matching index of the parameter data and its distribution. In addition, the fitting support device 1 estimates (2) the content of the survey conducted on the subject in order to determine the parameter data.
  • the target person When only one candidate parameter data is presented to the technician as in the past, the target person is made to evaluate whether the device set using the candidate parameter data is suitable for himself / herself. Then, when the subject evaluates that it conforms, the fitting is finished. However, in actual fitting, it is difficult to adapt the device to the target person with one fitting.
  • the technician must correct the parameters. However, in the case of a technician with a low skill level, he / she does not know how to correct the parameter data and repeats trial and error, so that the fitting time becomes long.
  • the matching index of the parameter data and its distribution are estimated, and the entire distribution and a plurality of promising candidate parameters are presented to the technician, so that the technician takes time for fitting. To shorten.
  • candidate parameters it is easy to adapt the parameters to the target person.
  • the question content, inspection content, etc. (survey content) necessary to obtain the necessary explanatory variables are requested and presented to the technician, the target person, or both. Subsequently, the response to the survey content, that is, the missing test information, the survey information, or both (explanatory variables) is obtained. After that, the technician causes the fitting support device 1 to estimate the parameter data again using the explanatory variables.
  • the fitting support device 1 has a selection unit 24 and an output information generation unit 25 in addition to the acquisition unit 2 and the estimation unit 3. Further, the estimation unit 3 has a learning model 26.
  • the selection unit 24 has a learning model 27 and a learning model 28. However, the learning models 26, 27, and 28 may be provided inside the fitting support device 1 or may be provided outside the fitting support device 1.
  • the input device 21 is a device used for inputting inspection information and survey information into the fitting support device 1. Specifically, the input device 21 first acquires inspection information and survey information of the target person from an information processing device or a storage device provided in a store, a manufacturer, a related facility, or the like of the device. Subsequently, the input device 21 transmits the acquired inspection information and the investigation information to the acquisition unit 2 of the fitting support device 1 by using communication such as wired or wireless.
  • the input device 21 supports the fitting of the acquired information when the inspection information used for fitting, the survey information, or both of them have insufficient information, but the insufficient information can be acquired thereafter. It is used when additional input is made to the device 1.
  • the missing information indicates the explanatory variables (variables of unobserved data) that have not yet been acquired when the explanatory variables necessary for obtaining the parameter data (objective variables) are known.
  • the input device 21 is, for example, an information processing device such as a personal computer, a mobile computer, a smartphone, or a tablet. Further, a plurality of input devices 21 may be prepared and inspection information and survey information may be input from separate input devices 21.
  • the output device 22 acquires the output information converted into an outputable format by the output information generation unit 25, and outputs the generated image, sound, and the like based on the output information.
  • the output device 22 is, for example, an image display device using a liquid crystal, an organic EL (Electro Luminescence), or a CRT (Cathode Ray Tube). Further, the image display device may include an audio output device such as a speaker.
  • the output device 22 may be a printing device such as a printer.
  • the fitting support device 1 will be described in detail.
  • the acquisition unit 2 acquires inspection information and investigation information from the input device 21 in the operation phase. Specifically, the acquisition unit 2 receives the inspection information that represents the result of the inspection performed on the target person and the survey information that represents the result of the survey related to the target person, which is transmitted from the input device 21. ..
  • the estimation unit 3 inputs the acquired inspection information and survey information, and estimates the goodness of fit index indicating the goodness of fit of the parameter data and its distribution. Further, when the inspection information used for fitting, the investigation information, or both of them have insufficient information, the estimation unit 3 acquires the missing information, adds the acquired information, and again. , Estimate the matching index of the parameter data and its distribution.
  • the estimation unit 3 first acquires inspection information and survey information (explanatory variables). Subsequently, the estimation unit 3 inputs the acquired inspection information and survey information (explanatory variable) into the learning model 26, and estimates the goodness of fit index representing the goodness of fit of the parameter data and its distribution.
  • the input can be expressed as Equation 1.
  • the distribution of the matching index of the parameter data can be expressed as Equation 2 using, for example, a probability distribution.
  • Equation 4 the probability distribution of the parameter data (objective variable) can be expressed as in Equation 4.
  • Equation 4 the unobserved data Z 3 and Z 4 are shown in the case where they do not depend on the observed data other than the unobserved data Z 3 and Z 4 , but they may depend on them.
  • the learning model 26 is set in the device in the past with the inspection information (explanatory variable) of a plurality of users acquired in the past, the survey information (explanatory variable) of the plurality of users acquired in the past, and the device. It is a learning model generated by machine learning with parameter data as input. Further, the learning model 26 inputs explanatory variables and outputs a matching index of each objective variable of the parameter data and its distribution. The details of the learning model 26 will be described later.
  • the learning model 26 is provided inside the fitting support device 1 in FIG. 2, it may be provided in an information processing device or a storage device (not shown in FIG. 2) provided outside the fitting support device 1. Good.
  • the estimation unit 3 has a configuration capable of communicating with the information processing device or the storage device by communication with each other.
  • the selection unit 24 is associated with the missing information in order to make up for the missing information when the inspection information used for fitting, the survey information, or both of them have missing information. Select the question information you asked.
  • the selection unit 24 first detects whether or not there is insufficient information in the inspection information used for fitting, the survey information, or both (explanatory variables).
  • the selection unit 24 sets a binary value indicating whether or not the information has been acquired for each explanatory variable (for example, information used to obtain the missing information). (Variables to be elements, etc.) are input to the learning model 27, and the probability distribution of the missing information is estimated.
  • the probability distribution (frequency of appearance in the learning data) of the missing information can be expressed as in Equation 5.
  • the learning model 27 is generated by machine learning by inputting the inspection information (explanatory variable) of a plurality of users acquired in the past and the survey information (explanatory variable) of a plurality of users acquired in the past in the learning phase. It is a learning model that is done. Further, the learning model 27 uses information (for example, a vector whose element is a binary value indicating whether or not the information has been acquired) used for obtaining the missing information for each inspection information and survey information (explanatory variable). Input and output the probability distribution of each unacquired explanatory variable. The details of the learning model 27 will be described later.
  • the selection unit 24 uses information to obtain the missing information (for example, a vector having a binary value as an element indicating whether or not the information has been acquired for each explanatory variable). Is input to the learning model 28, and the simultaneous distribution of the parameter data and the missing information is estimated.
  • the joint distribution can be expressed as in Equation 6.
  • the learning model 28 is set in the device in the past with the inspection information (explanatory variable) of a plurality of users acquired in the past, the survey information (explanatory variable) of the plurality of users acquired in the past, and the device. It is a learning model generated by machine learning with parameter data (objective variable) as input. Further, the learning model 28 inputs inspection information, survey information (explanatory variable), and parameter data (objective variable), and outputs a joint distribution of the explanatory variable and the objective variable. The details of the learning model 28 will be described later.
  • the selection unit 24 calculates the amount of information (importance) for each missing information by using the estimated probability distribution and the joint distribution.
  • the amount of information can be expressed as in Equation 7.
  • the selection unit 24 calculates the amount of information representing the correlation with the parameter data (objective variable) for each missing information, and based on this amount of information, one of the missing information. Select the above information.
  • the amount of information may be, for example, a conditional amount of information. It is desirable that the selection unit 24 selects the missing information in descending order of importance indicated by the amount of information.
  • the selection unit 24 selects the question information by referring to the selection information associated with the identification information that identifies the explanatory variable and the question information based on the selected missing information.
  • the question information is, for example, information representing the content of a question to be asked to the target person in order to acquire the missing information.
  • the selection information is stored in, for example, a storage device provided inside or outside the fitting support device 1.
  • FIG. 3 is a diagram showing an example of a data structure of selection information.
  • FIG. 3 shows an example of selection information when the device is a hearing aid.
  • the hearing aid collects sound using, for example, a sound collecting unit (microphone), amplifies and processes the sound collected by using the processing unit, and amplifies and processes using the output unit (receiver). It is a device that outputs sound.
  • the "item” stores the inspection information and the identification information that identifies the survey information
  • the "question content” is the question information indicating the inspection information and the question content used to acquire the survey information. Is remembered.
  • the output information generation unit 25 generates output information used to output the matching index of the parameter data estimated by the estimation unit 3 to the output device 22. Then, the output information generation unit 25 outputs the generated output information to the output device 22.
  • FIG. 4 is a diagram showing an example of the data structure of the parameter data. Further, FIG. 4 shows parameter data when the device is a hearing aid.
  • the parameter data is, for example, data used for adjusting one or more of the sound collecting unit, the processing unit, the output unit, or each of these units provided in the hearing aid.
  • each of the data (objective variables) of the parameter data has a frequency of each type of hearing aid and output level (50 [dB], 70 [dB], 90 [dB], maximum output level). Characteristic values, noise suppression strength, howling suppression strength, directivity type, impact sound suppression degree, etc. can be considered.
  • the parameter data includes "item” information representing the item of the objective variable possessed by the parameter data, "parameter” information representing the estimated parameter data (objective variable), and the type of data. It is associated with the information of the "data type" to be represented.
  • FIG. 5 is a diagram showing a display example of the distribution of the conformity index of the parameter data.
  • the objective variable is the output level 30 [dB], 40 [dB] corresponding to the frequencies 250 [Hz], 500 [Hz], respectively. ] ... Therefore, in order to assist in determining these objective variables, the estimated probability distribution for each objective variable is displayed. By displaying the probability distribution for each objective variable to be displayed in this way, candidates for a plurality of objective variables can be presented to the technician.
  • FIG. 6 is a diagram showing a display example of the distribution of the conformity index of the parameter data. Further, the probability distribution for each objective variable may be displayed using a column chart. Further, the magnitude of the probability may be displayed using a heat map.
  • the output information generation unit 25 generates output information used to output the question content represented by the question information selected by the selection unit 24 to the output device 22. Then, the output information generation unit 25 outputs the generated output information to the output device 22.
  • FIG. 7 is a diagram showing an example of a data structure of inspection information.
  • FIG. 8 is a diagram showing an example of the data structure of the survey information.
  • the examination information has at least one or more information of air conduction audiogram, bone conduction audiogram, discomfort threshold, and speech intelligibility.
  • the test information includes "item” information indicating the item of the hearing test, "hearing test result” information indicating the hearing test result, and "data type” indicating the data type. The information is associated and stored in the storage device.
  • the survey information is, for example, information on at least the subject's age, gender, occupation, whereabouts, family structure, medical history, treatment history, device usage history, symptoms, and physical characteristics (eg, height, weight). , Otoacoustic emission, etc.).
  • the ear acoustics are information representing the acoustic characteristics of the ear.
  • the survey information may include information representing the living environment sound of the subject, information representing the taste, and adjustment information.
  • Living environment sound is information representing the sound heard in the subject's daily life.
  • the taste is information expressing the taste of the subject for the sound.
  • the adjustment information is information indicating the adjustment date and time of the device, the adjustment place, the adjuster identification number, and the like.
  • the attribute information may include information indicating the type of hearing aid.
  • the adjustment information includes "item” information representing the survey (attribute, background, etc.) items related to the target person, and "survey content” information representing the survey content of the target person.
  • the information of "data type” indicating the data type is associated with and stored in the storage device.
  • the output information generation unit 25 generates output information used to output the amount of information calculated by the selection unit 24 to the output device 22. Then, the output information generation unit 25 outputs the generated output information to the output device 22.
  • FIG. 9 is a diagram showing a display example of the amount of information corresponding to the missing information. For example, it is conceivable to display a graph as shown in FIG. 9 showing the amount of information for the missing information.
  • FIG. 10 is a diagram showing an example of a system having a learning device.
  • the learning device 31 shown in FIG. 10 is a device that generates learning models 26, 27, and 28 using machine learning.
  • the system having the learning device 31 shown in FIG. 10 has a storage device 32 in addition to the learning device 31. Further, the learning device 31 has an acquisition unit 33, a classification unit 34, a classification unit 35, and a generation unit 36.
  • the storage device 32 stores explanatory variables (inspection information, survey information) acquired in the past and objective variables (parameter data).
  • the plurality of inspection information acquired in the past is information representing the results of inspections conducted by users of a plurality of devices in the past for each user.
  • the plurality of adjustment information acquired in the past is the adjustment information acquired by the users of a plurality of devices for each user in the past.
  • the parameter data acquired in the past is the parameter data used for adjusting the equipment in the fitting performed by the technician with a high skill level for the users of a plurality of equipment in the past. Further, the parameter data acquired in the past is the parameter data used for adjusting the equipment in the fitting performed by using the fitting support device 1 for the users of a plurality of equipment in the past.
  • the acquisition unit 33 acquires the learning data acquired in the past. Specifically, the acquisition unit 33 acquires learning data such as a plurality of inspection information, adjustment information, and parameter data acquired in the past from the storage device 32, and transmits the acquired learning data to the classification unit 34.
  • the classification unit 34 classifies the received learning data. Specifically, the classification unit 34 first compares a score (satisfaction level) indicating whether or not the user is satisfied with the fitting in the past with a preset threshold value. Subsequently, when the score is equal to or higher than the threshold value, the classification unit 34 classifies the learning data associated with the score equal to or higher than the threshold value.
  • a score satisfaction level
  • the classification unit 34 can make the learning models 26, 27, 28 train using only the learning data having a high degree of satisfaction, so that the parameter data estimated using the learning models 26, 27, 28 can be used. , User satisfaction is high.
  • the classification unit 35 further classifies the learning data classified by the classification unit 34. Specifically, the classification unit 35 further executes a clustering process on the learning data classified by the classification unit 34.
  • high-dimensional information for example, character string, time series signal, etc.
  • a deep neural network is converted into a vector value.
  • the converted vector value is converted into low-dimensional information (for example, a label) by using the k-nearest neighbor method.
  • unsupervised learning can be performed by using a clustering method such as the k-means method. You can make labels.
  • the learning models 26, 27, and 28 can be trained using only the learning data having a high degree of satisfaction, the parameter data estimated using the learning models 26, 27, 28. Will increase user satisfaction. Further, since the learning data can be reduced, the amount of computational resources can be reduced, such as shortening the learning time and reducing the memory usage.
  • the classification unit 35 by executing the clustering process on the learning data classified by the classification unit 34, the amount of computational resources can be further reduced in the case of learning.
  • the generation unit 36 performs machine learning using the learning data classified by the classification unit 35 to generate learning models 26, 27, 28, and stores the generated learning models 26, 27, 28 in the estimation unit 3.
  • the generation unit 36 may store the learning models 26, 27, and 28 in a system having the learning device 31, a system having the fitting support device 1, or a storage device other than the system.
  • machine learning can be, for example, supervised learning or semi-supervised learning.
  • learning such as regression (least squares method, Bayesian linear regression, regression method such as random forest), and multiclass classification (algorithm such as decision tree).
  • regression least squares method, Bayesian linear regression, regression method such as random forest
  • multiclass classification algorithm such as decision tree
  • the learning data classified by the classification unit 35 may be used, the learning data classified by the classification unit 34 may be used, or the unclassified learning data may be used. You may use it.
  • learning data used to generate the learning models 26, 27, 28 inspection information, adjustment information, and parameter data that have been determined to be in conformity in the past may be used.
  • the learning device 31 can generate learning models 26, 27, and 28 that take into account not only the inspection information acquired in the past but also the adjustment information and parameter data acquired in the past.
  • learning models 26, 27, and 28 can be generated by inputting the results of fitting performed by a technician with a high skill level.
  • FIG. 11 is a diagram showing an example of the operation of the fitting support device.
  • FIGS. 2 to 9 will be referred to as appropriate.
  • the fitting support method is implemented by operating the fitting support device. Therefore, the description of the fitting support method in the present embodiment will be replaced with the following description of the operation of the fitting support device.
  • FIG. 12 is a diagram showing an example of the operation of the learning device.
  • FIG. 10 will be referred to as appropriate.
  • the learning method is implemented by operating the learning device. Therefore, the description of the learning method in the present embodiment is replaced with the following description of the operation of the learning device.
  • the acquisition unit 2 acquires the inspection information and the investigation information (explanatory variable) of the target person from the input device 21 in the operation phase (step A1). Specifically, in step A1, the acquisition unit 2 receives the inspection information transmitted from the input device 21 indicating the result of the inspection performed on the target person and the survey information representing the result of the investigation related to the target person. And receive.
  • the estimation unit 3 uses the acquired inspection information and survey information as inputs to estimate the goodness-of-fit index representing the goodness of fit of the parameter data and its distribution (step A2). Specifically, in step A2, the estimation unit 3 first acquires inspection information and survey information (explanatory variables). Subsequently, in step A2, the estimation unit 3 inputs the acquired inspection information and survey information (explanatory variable) into the learning model 26, and estimates the goodness-of-fit index representing the goodness of fit of the parameter data and its distribution.
  • the selection unit 24 selects question information to be used to supplement the missing information when there is missing information in the inspection information, the survey information, or both of them used for fitting. (Step A3). Specifically, in step A3, the selection unit 24 first detects whether or not there is insufficient information in the inspection information used for fitting, the survey information, or both (explanatory variables).
  • step A3 when there is missing information (explanatory variable), the selection unit 24 determines whether or not the information used to obtain the missing information (for example, information has been acquired for each explanatory variable). (A vector having the represented binary value as an element, etc.) is input to the learning model 27, and the probability distribution of the missing information is estimated. Further, in step A3, when there is missing information, the selection unit 24 elements the information used to obtain the missing information (for example, a binary value indicating whether or not the information has been acquired for each explanatory variable). (Such as the vector to be used) is input to the learning model 28, and the simultaneous distribution of the parameter data and the missing information is estimated.
  • the information used to obtain the missing information for example, information has been acquired for each explanatory variable.
  • step A3 the selection unit 24 calculates the amount of information (importance) for each missing information by using the estimated probability distribution and the joint distribution. Subsequently, in step A3, the selection unit 24 calculates the amount of information representing the correlation with the parameter data (objective variable) for each missing information, and based on this amount of information, among the missing information. Select one or more pieces of information from.
  • the selection unit 24 refers to the selection information associated with the identification information for identifying the explanatory variable and the question information based on the selected missing information, and selects the question information.
  • the output information generation unit 25 generates output information used to output the matching index of the parameter data estimated by the estimation unit 3 to the output device 22 (step A4). Then, the output information generation unit 25 outputs the generated output information to the output device 22 (step A5).
  • the matching index of the parameter data is displayed as shown in FIGS. 5 and 6, for example.
  • the probability distribution for each objective variable may be displayed using a column chart.
  • the magnitude of the probability may be displayed by using a heat map, a contour map, or the like.
  • step A4 the output information generation unit 25 generates output information used to output the question content represented by the question information selected by the selection unit 24 to the output device 22. Then, in step A5, the output information generation unit 25 outputs the generated output information to the output device 22.
  • step A4 the output information generation unit 25 generates output information used to output the amount of information calculated by the selection unit 24 to the output device 22. Then, in step A5, the output information generation unit 25 outputs the generated output information to the output device 22.
  • steps A1 to A5 are repeated until the inspection information and the survey information (explanatory variables) necessary for determining the parameter data are prepared.
  • step A1 the missing information acquired by using the question content is added, and the processes of steps A2 to A5 are executed again.
  • the acquisition unit 33 acquires the learning data acquired in the past (step B1). Specifically, in step B1, the acquisition unit 33 acquires learning data such as a plurality of inspection information, survey information, and parameter data acquired in the past from the storage device 32, and transfers the acquired learning data to the classification unit 34. Send.
  • the classification unit 34 classifies the received learning data (step B2). Specifically, in step B2, the classification unit 34 first compares a score (satisfaction level) indicating whether or not the user is satisfied with the fitting in the past with a preset threshold value. Subsequently, in step B2, when the score is equal to or higher than the threshold value, the classification unit 34 classifies the learning data associated with the score equal to or higher than the threshold value.
  • a score evaluation level
  • the classification unit 35 further classifies the learning data classified in step B2 (step B3). Specifically, in step B3, the classification unit 35 further executes a clustering process on the learning data classified in step B2. In the clustering process, high-dimensional information is converted into low-dimensional information.
  • the generation unit 36 causes machine learning using the learning data classified by the classification unit 35 to generate learning models 26, 27, 28 (step B4), and generates the generated learning models 26, 27, 28. It is stored in the estimation unit 3 (step B5). Alternatively, in step B5, the generation unit 36 may store the learning models 26, 27, and 28 in a system having the learning device 31, a system having the fitting support device 1, or a storage device other than those.
  • the learning data classified by the classification unit 35 may be used, the learning data classified by the classification unit 34 may be used, or the unclassified learning data may be used. You may use it.
  • learning data used to generate the learning models 26, 27, 28 inspection information, adjustment information, and parameter data that have been determined to be in conformity in the past may be used.
  • the fitting support device 1 can estimate the matching index of the parameter data and its distribution, so that the matching index of the estimated parameter data and its distribution can be presented to the technician. Therefore, it is possible to present a plurality of candidate parameters to the technician.
  • the fitting support device 1 outputs a probability distribution to the output device as the distribution of the matching index of the parameter data, so that even a technician with a low skill level can refer to the output probability distribution and the target person. Appropriate fitting can be performed for. Therefore, the skill of a technician with a low skill level can be improved, and the accuracy of fitting can be improved.
  • a specific fitting policy can be presented, so that the technician can take the time required for fitting. Can be shortened. Further, by using the candidate parameters, the parameters can be easily adapted to the target person.
  • the specific fitting policy can be clarified. Therefore, the time required for the technician to perform the fitting can be shortened. In addition, since the necessary explanatory variables can be arranged as much as possible, the accuracy of estimating the parameters suitable for the target person can be improved.
  • the program for estimating the matching index of the parameter data and its distribution in the embodiment of the present invention may be a program that causes a computer to execute steps A1 to A5 shown in FIG. By installing this program on a computer and executing it, the fitting support device and the fitting support method according to the present embodiment can be realized.
  • the computer processor functions as an acquisition unit 2, an estimation unit 3, a selection unit 24, and an output information generation unit 25, and performs processing.
  • the program for estimating the matching index of the parameter data and its distribution in the present embodiment may be executed by a computer system constructed by a plurality of computers.
  • each computer may function as any of the acquisition unit 2, the estimation unit 3, the selection unit 24, and the output information generation unit 25, respectively.
  • the program for generating the learning model according to the embodiment of the present invention may be a program that causes a computer to execute steps B1 to B5 shown in FIG. By installing this program on a computer and executing it, the learning device and the learning method according to the present embodiment can be realized.
  • the computer processor functions as an acquisition unit 33, a classification unit 34, a classification unit 35, and a generation unit 36, and performs processing.
  • the program for generating the learning model in the present embodiment may be executed by a computer system constructed by a plurality of computers.
  • each computer may function as any of the acquisition unit 33, the classification unit 34, the classification unit 35, and the generation unit 36, respectively.
  • FIG. 13 is a block diagram showing an example of a computer that realizes the fitting support device or the learning device according to the embodiment of the present invention.
  • the computer 110 includes a CPU (Central Processing Unit) 111, a main memory 112, a storage device 113, an input interface 114, a display controller 115, a data reader / writer 116, and a communication interface 117. And. Each of these parts is connected to each other via a bus 121 so as to be capable of data communication.
  • the computer 110 may include a GPU (Graphics Processing Unit) or an FPGA (Field-Programmable Gate Array) in addition to the CPU 111 or in place of the CPU 111.
  • the CPU 111 expands the programs (codes) of the present embodiment stored in the storage device 113 into the main memory 112 and executes them in a predetermined order to perform various operations.
  • the main memory 112 is typically a volatile storage device such as a DRAM (Dynamic Random Access Memory).
  • the program according to the present embodiment is provided in a state of being stored in a computer-readable recording medium 120.
  • the program in the present embodiment may be distributed on the Internet connected via the communication interface 117.
  • the storage device 113 include a semiconductor storage device such as a flash memory in addition to a hard disk drive.
  • the input interface 114 mediates data transmission between the CPU 111 and an input device 118 such as a keyboard and mouse.
  • the display controller 115 is connected to the display device 119 and controls the display on the display device 119.
  • the data reader / writer 116 mediates the data transmission between the CPU 111 and the recording medium 120, reads the program from the recording medium 120, and writes the processing result in the computer 110 to the recording medium 120.
  • the communication interface 117 mediates data transmission between the CPU 111 and another computer.
  • the recording medium 120 include a general-purpose semiconductor storage device such as CF (CompactFlash (registered trademark)) and SD (SecureDigital), a magnetic recording medium such as a flexible disk, or a CD-.
  • CF CompactFlash (registered trademark)
  • SD Secure Digital
  • magnetic recording medium such as a flexible disk
  • CD- CompactDiskReadOnlyMemory
  • optical recording media such as ROM (CompactDiskReadOnlyMemory).
  • fitting support device 1 or the learning device 31 in the present embodiment can also be realized by using hardware corresponding to each part instead of the computer on which the program is installed. Further, the fitting support device 1 or the learning device 31 may be partially realized by a program and the rest may be realized by hardware.
  • Appendix 1 An acquisition unit that acquires inspection information representing the results of inspections conducted on the target person and survey information representing the results of the survey related to the target person. In the fitting of the parameter data used to fit the device to the target person, the acquired inspection information and the survey information are input to estimate the goodness of fit index representing the goodness of fit of the parameter data and its distribution.
  • Appendix 2 The fitting support device according to Appendix 1.
  • An output information generator that generates output information representing the matching index of the parameter data and its distribution and outputs it to the output device.
  • a fitting support device characterized by having.
  • the fitting support device described in Appendix 2 The estimation unit is generated by machine learning by inputting inspection information of a plurality of users acquired in the past, survey information of a plurality of users acquired in the past, and parameter data set in the device in the past.
  • the fitting support device described in Appendix 3 If the inspection information used for fitting, the survey information, or both of them have missing information, the question information associated with the missing information is selected to make up for the missing information. To do, the selection part, the output information generation unit is a fitting support device characterized in that the output information representing the question content is generated using the question information and output to the output device.
  • the fitting support device (Appendix 5) The fitting support device according to Appendix 4.
  • the selection unit calculates an amount of information representing the correlation between the deficient information and the parameter data with respect to the deficient information, and based on the calculated amount of information, the deficiency
  • step 7 A step of acquiring inspection information representing the results of an inspection conducted on a subject and survey information representing the results of a survey related to the subject.
  • step and A fitting support method characterized by having.
  • Appendix 8 The fitting support method described in Appendix 7 (C) A step of generating output information representing the conformity index of the parameter data and its distribution and outputting it to the output device.
  • Appendix 10 The fitting support method described in Appendix 9 (D) If the inspection information used for fitting, the survey information, or both of them have missing information, a question associated with the missing information to supplement the missing information. Select information, have steps, A fitting support method according to the step (c), wherein the output information representing the question content is generated using the question information and output to an output device.
  • step and A computer-readable recording medium that records a program, including instructions to execute.
  • Appendix 14 The computer-readable recording medium according to Appendix 13.
  • the program is on the computer (C) Further including an instruction to execute a step, which generates output information representing the conformity index of the parameter data and its distribution and outputs it to the output device.
  • (Appendix 15) The computer-readable recording medium according to Appendix 14.
  • the machine uses the inspection information of a plurality of users acquired in the past, the survey information of a plurality of users acquired in the past, and the parameter data set in the device in the past as inputs.
  • a computer-readable recording medium generated by learning that uses a learning model used for estimating the conformity index of the parameter data and its distribution.
  • Appendix 16 The computer-readable recording medium according to Appendix 15.
  • the program is on the computer (D) If the inspection information used for fitting, the survey information, or both of them have missing information, a question associated with the missing information to supplement the missing information. Includes additional instructions to select information, perform steps,
  • Appendix 18 A computer-readable recording medium according to Appendix 16 or 17.
  • the acquired information is added to the matching index of the parameter data and the matching index thereof.
  • a computer-readable recording medium characterized by estimating the distribution.
  • the accuracy of fitting can be improved.
  • the present invention is useful in a field where fitting such as a device worn on a living body to be adapted to a subject is required.
  • the device may be a hearing aid, an earphone, a smart glass, a head-mounted display, a smart watch, a musical instrument, a medical device, a nursing robot, or the like.
  • the device may be a device used without being attached to a living body.

Landscapes

  • Physics & Mathematics (AREA)
  • Engineering & Computer Science (AREA)
  • Acoustics & Sound (AREA)
  • Otolaryngology (AREA)
  • Neurosurgery (AREA)
  • General Health & Medical Sciences (AREA)
  • Health & Medical Sciences (AREA)
  • Signal Processing (AREA)
  • Artificial Intelligence (AREA)
  • Automation & Control Theory (AREA)
  • Evolutionary Computation (AREA)
  • Fuzzy Systems (AREA)
  • Mathematical Physics (AREA)
  • Software Systems (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

A fitting assistance device 1 comprises: an acquisition unit 2 for acquiring inspection information that represents the result of an inspection executed on a subject and examination information that represents the result of an examination pertaining to the subject; and an estimation unit 3 that, in parameter data fitting used to make an apparatus suit the subject, uses the acquired inspection information and examination information as inputs to estimate a suitability index that represents the suitability of parameter data and a distribution of the suitability index.

Description

フィッティング支援装置、フィッティング支援方法、及びコンピュータ読み取り可能な記録媒体Fitting support device, fitting support method, and computer-readable recording medium
　本発明は、フィッティング支援装置、フィッティング支援方法に関し、更には、これらを実現するためのプログラムを記録しているコンピュータ読み取り可能な記録媒体に関する。 The present invention relates to a fitting support device and a fitting support method, and further relates to a computer-readable recording medium that records a program for realizing these.
　生体に装着して用いる装着機器は、生体に合わせた調整（フィッティング）が必要である。例えば、補聴器の場合、技能者が、対象者の聴力、耳の構造、生活環境などを加味してフィッティングを実施している。 The mounting device used by mounting it on the living body needs to be adjusted (fitting) according to the living body. For example, in the case of a hearing aid, a technician performs fitting in consideration of the subject's hearing, ear structure, living environment, and the like.
　ところが、技能者の熟練度（技能レベル）によっては、対象者に対して適切なフィッティングが実施されないことがある。そこで、フィッティングを支援するシステムが提案されている。 However, depending on the skill level (skill level) of the technician, proper fitting may not be performed for the target person. Therefore, a system that supports fitting has been proposed.
　関連する技術として、特許文献１には、補聴器のフィッティングを自動で実施するフィッティング装置が開示されている。特許文献１に開示のフィッティング装置によれば、あらかじめ作成された環境音を対象者に試聴させ、試聴させた環境音の聞こえ方を対象者に評価させ、対象者が満足できるようになるまで補聴器の調整を繰り返すことにより、フィッティングをする装置である。 As a related technique, Patent Document 1 discloses a fitting device that automatically fits a hearing aid. According to the fitting device disclosed in Patent Document 1, the subject is made to audition the environmental sound created in advance, the subject is made to evaluate how to hear the auditioned environmental sound, and the hearing aid is used until the subject is satisfied. It is a device that fits by repeating the adjustment of.
特開２００１－００８２９５号公報Japanese Unexamined Patent Publication No. 2001-008295
　しかしながら、特許文献１に開示のフィッティング装置は、環境音に含まれる騒音をうるさく感じないようにするフィッティングを行うことができるが、熟練度の高い技能者が実施するようなフィッティングを実施することは難しい。 However, although the fitting device disclosed in Patent Document 1 can perform fitting so as not to feel the noise contained in the environmental sound noisy, it is not possible to carry out the fitting as performed by a highly skilled technician. difficult.
　また、装着機器に対してフィッティングを行う場合には、様々な対象者がいるので、対象者それぞれに対して特徴を加味する必要がある。そのため、熟練度の低い技能者では、対象者に対して適切なフィッティングを実施することは難しい。 In addition, when fitting to the attached device, there are various target persons, so it is necessary to add characteristics to each target person. Therefore, it is difficult for a technician with a low skill level to perform appropriate fitting for the target person.
　本発明の目的の一例は、フィッティングの精度を向上させるフィッティング支援装置、フィッティング支援方法、及びコンピュータ読み取り可能な記録媒体を提供することにある。 An example of an object of the present invention is to provide a fitting support device, a fitting support method, and a computer-readable recording medium for improving the accuracy of fitting.
　上記目的を達成するため、本発明の一側面におけるフィッティング支援装置は、
　対象者に対して実施した検査の結果を表す検査情報と、前記対象者に関係する調査の結果を表す調査情報とを取得する、取得手段と、
　前記対象者に機器を適合させるために用いるパラメータデータのフィッティングにおいて、取得した前記検査情報と前記調査情報とを入力して、前記パラメータデータの適合度を表す適合指標及びその分布を推定する、推定手段と、
　を有することを特徴とする。 In order to achieve the above object, the fitting support device in one aspect of the present invention is
An acquisition means for acquiring inspection information representing the results of an inspection conducted on a subject and survey information representing the results of a survey related to the subject.
In the fitting of the parameter data used to fit the device to the target person, the acquired inspection information and the survey information are input to estimate the goodness of fit index representing the goodness of fit of the parameter data and its distribution. Means and
It is characterized by having.
　また、上記目的を達成するため、本発明の一側面におけるフィッティング支援方法は、
（ａ）対象者に対して実施した検査の結果を表す検査情報と、前記対象者に関係する調査の結果を表す調査情報とを取得し、
（ｂ）前記対象者に機器を適合させるために用いるパラメータデータのフィッティングにおいて、取得した前記検査情報と前記調査情報とを入力して、前記パラメータデータの適合度を表す適合指標及びその分布を推定する
　することを特徴とする。 Further, in order to achieve the above object, the fitting support method in one aspect of the present invention is:
(A) Obtain the inspection information indicating the result of the inspection conducted on the subject and the survey information representing the result of the investigation related to the subject.
(B) In fitting the parameter data used to fit the device to the target person, the acquired inspection information and the survey information are input to estimate the goodness of fit index representing the goodness of fit of the parameter data and its distribution. It is characterized by doing.
　更に、上記目的を達成するため、本発明の一側面におけるプログラムを記録したコンピュータ読み取り可能な記録媒体は、
　コンピュータに、
（ａ）対象者に対して実施した検査の結果を表す検査情報と、前記対象者に関係する調査の結果を表す調査情報とを取得する、ステップと、
（ｂ）前記対象者に機器を適合させるために用いるパラメータデータのフィッティングにおいて、取得した前記検査情報と前記調査情報とを入力して、前記パラメータデータの適合度を表す適合指標及びその分布を推定する、ステップと、
　を実行させる命令を含む、プログラムを記録していることを特徴とする。 Further, in order to achieve the above object, a computer-readable recording medium on which a program according to one aspect of the present invention is recorded is
On the computer
(A) A step of acquiring inspection information representing the results of an inspection conducted on a subject and survey information representing the results of a survey related to the subject.
(B) In fitting the parameter data used to fit the device to the target person, the acquired inspection information and the survey information are input to estimate the goodness of fit index representing the goodness of fit of the parameter data and its distribution. To do, step and
It is characterized by recording a program including an instruction to execute.
　以上のように本発明によれば、フィッティングの精度を向上させることができる。 As described above, according to the present invention, the accuracy of fitting can be improved.
図１は、フィッティング支援装置の一例を示す図である。FIG. 1 is a diagram showing an example of a fitting support device. 図２は、フィッティング支援装置を有するシステムの一例を示す図である。FIG. 2 is a diagram showing an example of a system having a fitting support device. 図３は、選択情報のデータ構造の一例を示す図である。FIG. 3 is a diagram showing an example of a data structure of selection information. 図４は、パラメータデータのデータ構造の一例を示す図である。FIG. 4 is a diagram showing an example of a data structure of parameter data. 図５は、パラメータデータの適合指標の表示例を示す図である。FIG. 5 is a diagram showing a display example of a conformity index of parameter data. 図６は、パラメータデータの適合指標の表示例を示す図である。FIG. 6 is a diagram showing a display example of a conformity index of parameter data. 図７は、検査情報のデータ構造の一例を示す図であるFIG. 7 is a diagram showing an example of a data structure of inspection information. 図８は、調査情報のデータ構造の一例を示す図である。FIG. 8 is a diagram showing an example of the data structure of the survey information. 図９は、不足している情報に対応する情報量の表示例を示す図である。FIG. 9 is a diagram showing a display example of the amount of information corresponding to the missing information. 図１０は、学習装置を有するシステムの一例を示す図である。FIG. 10 is a diagram showing an example of a system having a learning device. 図１１は、フィッティング支援装置の動作の一例を示す図である。FIG. 11 is a diagram showing an example of the operation of the fitting support device. 図１２は、学習装置の動作の一例を示す図である。FIG. 12 is a diagram showing an example of the operation of the learning device. 図１３は、フィッティング支援装置を実現するコンピュータの一例を示す図である。FIG. 13 is a diagram showing an example of a computer that realizes a fitting support device.
（実施の形態）
　以下、本発明の実施の形態について、図１から図１３を参照しながら説明する。 (Embodiment)
Hereinafter, embodiments of the present invention will be described with reference to FIGS. 1 to 13.
［装置構成］
　最初に、図１を用いて、本実施の形態におけるフィッティング支援装置１の構成について説明する。図１は、フィッティング支援装置の一例を示す図である。 [Device configuration]
First, the configuration of the fitting support device 1 according to the present embodiment will be described with reference to FIG. FIG. 1 is a diagram showing an example of a fitting support device.
　図１に示すフィッティング支援装置は、対象者に機器を適合させる調整（フィッティング）の精度を向上させる装置である。また、図１に示すように、フィッティング支援装置１は、取得部２と、推定部３とを有する。 The fitting support device shown in FIG. 1 is a device that improves the accuracy of adjustment (fitting) for adapting the device to the target person. Further, as shown in FIG. 1, the fitting support device 1 has an acquisition unit 2 and an estimation unit 3.
　機器とは、例えば、生体に装着して用いる機器などが考えられる。具体的には、機器として、補聴器、イヤホン、スマートグラス、ヘッドマウントディスプレイ、スマートウォッチ、楽器、医療用装具、介護ロボットなどが考えられる。ただし、機器は、生体に装着せずに用いる機器でもよい。 The device may be, for example, a device that is attached to a living body and used. Specifically, as devices, hearing aids, earphones, smart glasses, head-mounted displays, smart watches, musical instruments, medical devices, nursing robots, and the like can be considered. However, the device may be a device used without being attached to a living body.
　このうち、取得部２は、対象者に対して実施した検査の結果を表す検査情報と、対象者に関係する調査の結果を表す調査情報とを取得する。推定部３は、対象者に機器を適合させるために用いるパラメータデータのフィッティングにおいて、取得した検査情報と調査情報とを入力として、パラメータデータの適合度を表す適合指標及びその分布を推定する。 Of these, the acquisition unit 2 acquires inspection information representing the results of the inspection conducted on the target person and survey information representing the results of the survey related to the target person. In the fitting of the parameter data used to fit the device to the target person, the estimation unit 3 uses the acquired inspection information and the survey information as inputs to estimate the goodness of fit index representing the goodness of fit of the parameter data and its distribution.
　検査情報は、対象者に対して実施した検査の結果を表す情報（説明変数）である。調査情報は、対象者に関係する調査の結果を表す情報（説明変数）である。パラメータデータ（目的変数）は、対象とする機器のハードウェア、ソフトウェアを調整するために用いるデータ（設定値）である。 The inspection information is information (explanatory variable) representing the result of the inspection conducted on the subject. The survey information is information (explanatory variable) representing the results of the survey related to the subject. Parameter data (objective variable) is data (set value) used to adjust the hardware and software of the target device.
　適合指標は、フィッティングにおけるパラメータデータの適合度を表す指標である。適合指標としては、例えば、パラメータデータの有望度を表す確率などが考えられる。他にも、ドメイン知識に基づいて定められているフィッティングに関する安全基準などの制約条件を満たしている数などを、適合指標として用いてもよい。 The goodness of fit index is an index showing the goodness of fit of parameter data in fitting. As the conformity index, for example, the probability of expressing the promising degree of the parameter data can be considered. In addition, a number that meets constraints such as safety standards for fittings determined based on domain knowledge may be used as a conformity index.
　推定部３は、過去に取得した複数の検査情報と、過去に取得した複数の調査情報と、過去に機器に設定されたパラメータデータとを入力として、機械学習により生成された、パラメータデータの適合指標及びその分布の推定に用いる学習モデルを有する。 The estimation unit 3 inputs a plurality of inspection information acquired in the past, a plurality of survey information acquired in the past, and parameter data set in the device in the past, and matches the parameter data generated by machine learning. It has a learning model used to estimate the index and its distribution.
　機械学習は、例えば、教師あり学習、半教師あり学習などが考えられる。具体的には、機械学習は、回帰（ベイズ線型回帰などの回帰手法）、多クラス分類（決定木などのアルゴリズム）などである。 Machine learning can be, for example, supervised learning or semi-supervised learning. Specifically, machine learning includes regression (regression method such as Bayesian linear regression), multiclass classification (algorithm such as decision tree), and the like.
　このように、本実施の形態において、フィッティング支援装置１は、パラメータデータの適合指標を推定できるので、推定したパラメータデータの適合指標及びその分布を技能者に提示できる。そのため、技能者は、対象者に対して適切なフィッティングを実施できる。 As described above, in the present embodiment, the fitting support device 1 can estimate the matching index of the parameter data, so that the matching index of the estimated parameter data and its distribution can be presented to the technician. Therefore, the technician can carry out appropriate fitting for the target person.
　具体的には、フィッティング支援装置１は、パラメータデータの適合指標の分布として確率分布を出力装置に出力することで、技能レベルの低い技能者でも、出力された確率分布を参考にして、対象者に対して適切なフィッティングを実施することができる。そのため、技能レベルの低い技能者の技能を向上させることができるので、フィッティングの精度を向上させることができる。 Specifically, the fitting support device 1 outputs a probability distribution to the output device as the distribution of the matching index of the parameter data, so that even a technician with a low skill level can refer to the output probability distribution and the target person. Appropriate fitting can be performed for. Therefore, the skill of a technician with a low skill level can be improved, and the accuracy of fitting can be improved.
［システム構成］
　続いて、図２を用いて、本実施の形態におけるフィッティング支援装置１の構成をより具体的に説明する。図２は、フィッティング支援装置１を有するシステムの一例を示す図である。 [System configuration]
Subsequently, the configuration of the fitting support device 1 according to the present embodiment will be described more specifically with reference to FIG. FIG. 2 is a diagram showing an example of a system having a fitting support device 1.
　図２に示すように、本実施の形態におけるフィッティング支援装置１を有するシステムは、フィッティング支援装置１に加え、入力装置２１と、出力装置２２とを有する。ただし、システムは、図２に示した構成に限定されるものではない。 As shown in FIG. 2, the system having the fitting support device 1 in the present embodiment has an input device 21 and an output device 22 in addition to the fitting support device 1. However, the system is not limited to the configuration shown in FIG.
　フィッティング支援装置１は、（１）パラメータデータの適合指標及びその分布を推定する。また、フィッティング支援装置１は、（２）パラメータデータを決定するために対象者に対して行う調査内容を推定する。 The fitting support device 1 estimates (1) the matching index of the parameter data and its distribution. In addition, the fitting support device 1 estimates (2) the content of the survey conducted on the subject in order to determine the parameter data.
（１）パラメータデータの適合指標及びその分布の推定について
　（１）においては、従来のように候補となるパラメータデータを技能者に一つだけ提示するのではなく、適合指標及びその分布までも推定することで、分布全体や有望な複数の候補となるパラメータを提示する。 (1) Estimating the matching index and its distribution of parameter data In (1), instead of presenting only one candidate parameter data to the technician as in the past, the matching index and its distribution are also estimated. By doing so, the parameters of the entire distribution and a plurality of promising candidates are presented.
　従来のように候補となるパラメータデータを技能者に一つだけ提示する場合、対象者に、その候補となるパラメータデータを用いて設定された機器が、自身に適合しているかを評価させる。そして、適合していると対象者が評価した場合、フィッティングを終了する。しかし、実際のフィッティングにおいて、一度のフィッティングで、機器を対象者に適合させることは困難である。 When only one candidate parameter data is presented to the technician as in the past, the target person is made to evaluate whether the device set using the candidate parameter data is suitable for himself / herself. Then, when the subject evaluates that it conforms, the fitting is finished. However, in actual fitting, it is difficult to adapt the device to the target person with one fitting.
　対象者が適合していないと評価した場合、技能者は、パラメータを修正しなくてはならない。ところが、技能レベルの低い技能者の場合、パラメータデータをどのように修正したらいいのかが分からず、試行錯誤を繰り返すため、フィッティングに要する時間が長くなる。 If the target person evaluates that it is not suitable, the technician must correct the parameters. However, in the case of a technician with a low skill level, he / she does not know how to correct the parameter data and repeats trial and error, so that the fitting time becomes long.
　つまり、具体的なフィッティングの方針が不明であるため、フィッティングに要する時間が長くなる。更に、最終的に決定したパラメータを用いて機器を調整しても、実際には対象者に適合していないこともある。 In other words, since the specific fitting policy is unknown, the time required for fitting will be long. Furthermore, even if the device is adjusted using the finally determined parameters, it may not actually be suitable for the subject.
　そこで、（１）においては、パラメータデータの適合指標及びその分布までも推定し、分布全体や有望な複数の候補となるパラメータを技能者に提示することで、技能者がフィッティングのために要する時間を短縮する。また、候補となるパラメータを用いることで、パラメータを対象者に適合し易くする。 Therefore, in (1), the matching index of the parameter data and its distribution are estimated, and the entire distribution and a plurality of promising candidate parameters are presented to the technician, so that the technician takes time for fitting. To shorten. In addition, by using candidate parameters, it is easy to adapt the parameters to the target person.
（２）パラメータデータを決定するために対象者に対して行う調査内容の推定について
　フィッティングにおいて、対象者に機器を適合させるためには、パラメータデータ（目的変数）を決定するために、ある程度の検査情報、調査情報などの説明変数が必要である。ところが、フィッティングにおいて、必要な検査情報、調査情報などの説明変数が揃わない（不足する）場合がある。 (2) Estimating the content of the survey conducted on the subject to determine the parameter data In fitting, in order to adapt the device to the subject, some inspection is performed to determine the parameter data (objective variable). Explanatory variables such as information and survey information are required. However, in fitting, there are cases where necessary explanatory variables such as inspection information and survey information are not available (insufficient).
　そうすると、説明変数が不足しているため、決定したパラメータデータが、実際には対象者に適合していない。そこで、（２）においては、必要な説明変数を得るために調査をし、調査結果に基づいて必要な説明変数を推定する。 Then, because the explanatory variables are insufficient, the determined parameter data does not actually match the target person. Therefore, in (2), a survey is conducted to obtain the necessary explanatory variables, and the necessary explanatory variables are estimated based on the survey results.
　具体的には、（２）においては、必要な説明変数を得るために必要な質問内容・検査内容など（調査内容）を求めて、技能者、又は対象者、又は両者に提示する。続いて、調査内容に対する応答、すなわち不足している検査情報、又は調査情報、又はそれら両方（説明変数）を得る。その後、技能者は、当該説明変数を用いて、再度、フィッティング支援装置１にパラメータデータを推定させる。 Specifically, in (2), the question content, inspection content, etc. (survey content) necessary to obtain the necessary explanatory variables are requested and presented to the technician, the target person, or both. Subsequently, the response to the survey content, that is, the missing test information, the survey information, or both (explanatory variables) is obtained. After that, the technician causes the fitting support device 1 to estimate the parameter data again using the explanatory variables.
　このように、（２）によれば、調査内容を提示できるので、具体的なフィッティングの方針が明確にできるため、技能者がフィッティングのために要する時間を短縮できる。また、必要な説明変数をできるだけ揃えることができるので、対象者に適合したパラメータを推定する精度を向上させることができる。 In this way, according to (2), since the survey contents can be presented, the specific fitting policy can be clarified, and the time required for the technician to perform the fitting can be shortened. In addition, since the necessary explanatory variables can be arranged as much as possible, the accuracy of estimating the parameters suitable for the target person can be improved.
　なお、フィッティング支援装置１は、取得部２、推定部３に加えて、選択部２４と、出力情報生成部２５とを有する。更に、推定部３は、学習モデル２６を有する。選択部２４は、学習モデル２７、学習モデル２８を有する。ただし、学習モデル２６、２７、２８は、フィッティング支援装置１の内部に設けてもよいし、フィッティング支援装置１の外部に設けてもよい。 The fitting support device 1 has a selection unit 24 and an output information generation unit 25 in addition to the acquisition unit 2 and the estimation unit 3. Further, the estimation unit 3 has a learning model 26. The selection unit 24 has a learning model 27 and a learning model 28. However, the learning models 26, 27, and 28 may be provided inside the fitting support device 1 or may be provided outside the fitting support device 1.
　入力装置２１は、検査情報及び調査情報を、フィッティング支援装置１に入力するために用いる装置である。具体的には、入力装置２１は、まず、機器の販売店、製造元、関連施設などに設けられている情報処理装置又は記憶装置などから、対象者の検査情報、調査情報を取得する。続いて、入力装置２１は、取得した検査情報及び調査情報を、フィッティング支援装置１の取得部２へ有線又は無線などの通信を用いて送信する。 The input device 21 is a device used for inputting inspection information and survey information into the fitting support device 1. Specifically, the input device 21 first acquires inspection information and survey information of the target person from an information processing device or a storage device provided in a store, a manufacturer, a related facility, or the like of the device. Subsequently, the input device 21 transmits the acquired inspection information and the investigation information to the acquisition unit 2 of the fitting support device 1 by using communication such as wired or wireless.
　また、入力装置２１は、フィッティングに用いる検査情報、又は調査情報、又はそれら両方に不足している情報があったが、その後に不足している情報が取得できた場合、取得した情報をフィッティング支援装置１へ追加入力する際に用いる。 Further, the input device 21 supports the fitting of the acquired information when the inspection information used for fitting, the survey information, or both of them have insufficient information, but the insufficient information can be acquired thereafter. It is used when additional input is made to the device 1.
　不足している情報とは、パラメータデータ（目的変数）を得るために必要な説明変数が分かっている場合に、まだ取得できていない説明変数（未観測データの変数）のことを示す。 The missing information indicates the explanatory variables (variables of unobserved data) that have not yet been acquired when the explanatory variables necessary for obtaining the parameter data (objective variables) are known.
　なお、入力装置２１は、例えば、パーソナルコンピュータ、モバイルコンピュータ、スマートフォン、タブレットなどの情報処理装置である。また、入力装置２１を複数用意して、別々の入力装置２１から検査情報及び調査情報を入力してもよい。 The input device 21 is, for example, an information processing device such as a personal computer, a mobile computer, a smartphone, or a tablet. Further, a plurality of input devices 21 may be prepared and inspection information and survey information may be input from separate input devices 21.
　出力装置２２は、出力情報生成部２５により、出力可能な形式に変換された、出力情報を取得し、その出力情報に基づいて、生成した画像及び音声などを出力する。出力装置２２は、例えば、液晶、有機ＥＬ（Electro Luminescence）、ＣＲＴ（Cathode Ray Tube）を用いた画像表示装置などである。更に、画像表示装置は、スピーカなどの音声出力装置などを備えていてもよい。なお、出力装置２２は、プリンタなどの印刷装置でもよい。 The output device 22 acquires the output information converted into an outputable format by the output information generation unit 25, and outputs the generated image, sound, and the like based on the output information. The output device 22 is, for example, an image display device using a liquid crystal, an organic EL (Electro Luminescence), or a CRT (Cathode Ray Tube). Further, the image display device may include an audio output device such as a speaker. The output device 22 may be a printing device such as a printer.
　フィッティング支援装置１について詳細に説明する。
　取得部２は、運用フェーズにおいて、入力装置２１から検査情報、調査情報を取得する。具体的には、取得部２は、入力装置２１から送信された、対象者に対して実施した検査の結果を表す検査情報と、対象者に関係する調査の結果を表す調査情報とを受信する。 The fitting support device 1 will be described in detail.
The acquisition unit 2 acquires inspection information and investigation information from the input device 21 in the operation phase. Specifically, the acquisition unit 2 receives the inspection information that represents the result of the inspection performed on the target person and the survey information that represents the result of the survey related to the target person, which is transmitted from the input device 21. ..
　推定部３は、運用フェーズにおいて、取得した検査情報及び調査情報を入力として、パラメータデータの適合度を表す適合指標及びその分布を推定する。更に、推定部３は、フィッティングに用いる検査情報、又は調査情報、又はそれら両方に不足している情報があった場合、不足している情報を取得した後、取得した情報を追加して、再度、パラメータデータの適合指標及びその分布を推定する。 In the operation phase, the estimation unit 3 inputs the acquired inspection information and survey information, and estimates the goodness of fit index indicating the goodness of fit of the parameter data and its distribution. Further, when the inspection information used for fitting, the investigation information, or both of them have insufficient information, the estimation unit 3 acquires the missing information, adds the acquired information, and again. , Estimate the matching index of the parameter data and its distribution.
　具体的には、推定部３は、まず、検査情報及び調査情報（説明変数）を取得する。続いて、推定部３は、取得した検査情報及び調査情報（説明変数）を、学習モデル２６に入力し、パラメータデータの適合度を表す適合指標及びその分布を推定する。 Specifically, the estimation unit 3 first acquires inspection information and survey information (explanatory variables). Subsequently, the estimation unit 3 inputs the acquired inspection information and survey information (explanatory variable) into the learning model 26, and estimates the goodness of fit index representing the goodness of fit of the parameter data and its distribution.
　例えば、学習モデル２６に入力する説明変数がすべて揃っている場合、入力は数１のように表すことができる。 For example, when all the explanatory variables to be input to the learning model 26 are prepared, the input can be expressed as Equation 1.
［数１］
Figure JPOXMLDOC01-appb-I000001
[Number 1]
Figure JPOXMLDOC01-appb-I000001
　そして、入力する説明変数がすべて揃っている場合には、パラメータデータ（目的変数）の適合指標の分布は、例えば、確率分布を用いて、数２のように表すことができる。 Then, when all the explanatory variables to be input are prepared, the distribution of the matching index of the parameter data (objective variable) can be expressed as Equation 2 using, for example, a probability distribution.
［数２］
Figure JPOXMLDOC01-appb-I000002
　また、学習モデル２６へ入力する説明変数が不足している場合、入力は数３のように表すことができる。数３の例では、不足している情報を未観測データＺ３、Ｚ４として表している。 [Number 2]
Figure JPOXMLDOC01-appb-I000002
Further, when the explanatory variables to be input to the learning model 26 are insufficient, the input can be expressed as in Equation 3. In the example of Equation 3, the missing information is represented as unobserved data Z 3 and Z 4 .
［数３］
Figure JPOXMLDOC01-appb-I000003
[Number 3]
Figure JPOXMLDOC01-appb-I000003
　また、未観測データＺ３、Ｚ４がある場合には、パラメータデータ（目的変数）の確率分布は、数４のように表すことができる。 Further, when there are unobserved data Z 3 and Z 4 , the probability distribution of the parameter data (objective variable) can be expressed as in Equation 4.
［数４］
Figure JPOXMLDOC01-appb-I000004
[Number 4]
Figure JPOXMLDOC01-appb-I000004
　ただし、数４では、未観測データＺ３、Ｚ４は、未観測データＺ３、Ｚ４以外の観測データに依存しない場合について示したが、依存してもよい。 However, in Equation 4, the unobserved data Z 3 and Z 4 are shown in the case where they do not depend on the observed data other than the unobserved data Z 3 and Z 4 , but they may depend on them.
　学習モデル２６は、学習フェーズにおいて、過去に取得した複数の利用者の検査情報（説明変数）と、過去に取得した複数の利用者の調査情報（説明変数）と、過去に機器に設定されたパラメータデータとを入力として、機械学習により生成される、学習モデルである。また、学習モデル２６は、説明変数を入力して、パラメータデータが有する目的変数それぞれの適合指標やその分布を出力する。学習モデル２６の詳細については後述する。 In the learning phase, the learning model 26 is set in the device in the past with the inspection information (explanatory variable) of a plurality of users acquired in the past, the survey information (explanatory variable) of the plurality of users acquired in the past, and the device. It is a learning model generated by machine learning with parameter data as input. Further, the learning model 26 inputs explanatory variables and outputs a matching index of each objective variable of the parameter data and its distribution. The details of the learning model 26 will be described later.
　なお、図２において学習モデル２６は、フィッティング支援装置１の内部に設けられているが、フィッティング支援装置１の外部に設けられた、図２に不図示の情報処理装置又は記憶装置に設けてもよい。その場合、推定部３は、情報処理装置又は記憶装置と、互いに通信によりやり取りが可能な構成とする。 Although the learning model 26 is provided inside the fitting support device 1 in FIG. 2, it may be provided in an information processing device or a storage device (not shown in FIG. 2) provided outside the fitting support device 1. Good. In that case, the estimation unit 3 has a configuration capable of communicating with the information processing device or the storage device by communication with each other.
　選択部２４は、運用フェーズにおいて、フィッティングに用いる検査情報、又は調査情報、又はそれら両方に不足している情報がある場合、不足している情報を補うために、不足している情報に関連付けられた質問情報を選択する。 In the operation phase, the selection unit 24 is associated with the missing information in order to make up for the missing information when the inspection information used for fitting, the survey information, or both of them have missing information. Select the question information you asked.
　具体的には、選択部２４は、まず、フィッティングに用いる検査情報、又は調査情報、又はそれら両方（説明変数）に、不足している情報があるか否かを検出する。 Specifically, the selection unit 24 first detects whether or not there is insufficient information in the inspection information used for fitting, the survey information, or both (explanatory variables).
　続いて、選択部２４は、不足している情報（説明変数）がある場合、不足している情報を得るために用いる情報（例えば、説明変数ごとに情報取得済みか否かを表すバイナリ値を要素とするベクトルなど）を、学習モデル２７に入力し、不足している情報の確率分布を推定する。不足している情報の確率分布（学習データにおける出現頻度）は、数５のように表すことができる。 Subsequently, when there is missing information (explanatory variable), the selection unit 24 sets a binary value indicating whether or not the information has been acquired for each explanatory variable (for example, information used to obtain the missing information). (Variables to be elements, etc.) are input to the learning model 27, and the probability distribution of the missing information is estimated. The probability distribution (frequency of appearance in the learning data) of the missing information can be expressed as in Equation 5.
［数５］
Figure JPOXMLDOC01-appb-I000005
[Number 5]
Figure JPOXMLDOC01-appb-I000005
　学習モデル２７は、学習フェーズにおいて、過去に取得した複数の利用者の検査情報（説明変数）と、過去に取得した複数の利用者の調査情報（説明変数）とを入力として、機械学習により生成される、学習モデルである。また、学習モデル２７は、検査情報及び調査情報（説明変数）ごとに不足している情報を得るために用いる情報（例えば、情報取得済みか否かを表すバイナリ値を要素とするベクトルなど）を入力して、未取得の説明変数それぞれの確率分布を出力する。学習モデル２７の詳細については後述する。 The learning model 27 is generated by machine learning by inputting the inspection information (explanatory variable) of a plurality of users acquired in the past and the survey information (explanatory variable) of a plurality of users acquired in the past in the learning phase. It is a learning model that is done. Further, the learning model 27 uses information (for example, a vector whose element is a binary value indicating whether or not the information has been acquired) used for obtaining the missing information for each inspection information and survey information (explanatory variable). Input and output the probability distribution of each unacquired explanatory variable. The details of the learning model 27 will be described later.
　また、選択部２４は、不足している情報がある場合、不足している情報を得るために用いる情報（例えば、説明変数ごとに情報取得済みか否かを表すバイナリ値を要素とするベクトル）を、学習モデル２８に入力し、パラメータデータと不足している情報との同時分布を推定する。同時分布は、数６のように表すことができる。 Further, when there is missing information, the selection unit 24 uses information to obtain the missing information (for example, a vector having a binary value as an element indicating whether or not the information has been acquired for each explanatory variable). Is input to the learning model 28, and the simultaneous distribution of the parameter data and the missing information is estimated. The joint distribution can be expressed as in Equation 6.
［数６］
Figure JPOXMLDOC01-appb-I000006
[Number 6]
Figure JPOXMLDOC01-appb-I000006
　学習モデル２８は、学習フェーズにおいて、過去に取得した複数の利用者の検査情報（説明変数）と、過去に取得した複数の利用者の調査情報（説明変数）と、過去に機器に設定されたパラメータデータ（目的変数）とを入力として、機械学習により生成される、学習モデルである。また、学習モデル２８は、検査情報及び調査情報（説明変数）とパラメータデータ（目的変数）を入力して、説明変数と目的変数の同時分布を出力する。学習モデル２８の詳細については後述する。 In the learning phase, the learning model 28 is set in the device in the past with the inspection information (explanatory variable) of a plurality of users acquired in the past, the survey information (explanatory variable) of the plurality of users acquired in the past, and the device. It is a learning model generated by machine learning with parameter data (objective variable) as input. Further, the learning model 28 inputs inspection information, survey information (explanatory variable), and parameter data (objective variable), and outputs a joint distribution of the explanatory variable and the objective variable. The details of the learning model 28 will be described later.
　続いて、選択部２４は、不足している情報ごとに、推定した確率分布と、同時分布とを用いて、情報量（重要度）を算出する。情報量は、数７のように表すことができる。 Subsequently, the selection unit 24 calculates the amount of information (importance) for each missing information by using the estimated probability distribution and the joint distribution. The amount of information can be expressed as in Equation 7.
［数７］
Figure JPOXMLDOC01-appb-I000007
[Number 7]
Figure JPOXMLDOC01-appb-I000007
　続いて、選択部２４は、不足している情報ごとに、パラメータデータ（目的変数）との相関を表す情報量を算出し、この情報量に基づいて、不足している情報の中から一つ以上の情報を選択する。情報量は、例えば、条件付き情報量などが考えられる。なお、選択部２４は、情報量の示す重要度が高い順に、不足している情報を選択することが望ましい。 Subsequently, the selection unit 24 calculates the amount of information representing the correlation with the parameter data (objective variable) for each missing information, and based on this amount of information, one of the missing information. Select the above information. The amount of information may be, for example, a conditional amount of information. It is desirable that the selection unit 24 selects the missing information in descending order of importance indicated by the amount of information.
　続いて、選択部２４は、選択した不足している情報に基づいて、説明変数を識別する識別情報と質問情報とが関連付けられた選択情報を参照し、質問情報を選択する。質問情報は、例えば、不足している情報を取得するために、対象者に対して実施する質問の内容を表す情報である。選択情報は、例えば、フィッティング支援装置１の内部又は外部に設けられた記憶装置に記憶されている。 Subsequently, the selection unit 24 selects the question information by referring to the selection information associated with the identification information that identifies the explanatory variable and the question information based on the selected missing information. The question information is, for example, information representing the content of a question to be asked to the target person in order to acquire the missing information. The selection information is stored in, for example, a storage device provided inside or outside the fitting support device 1.
　図３は、選択情報のデータ構造の一例を示す図である。図３には、機器を補聴器とした場合の選択情報の例が示されている。ここで、補聴器は、例えば、集音部（マイク）を用いて集音し、処理部を用いて集音された音を増幅及び加工し、出力部（レシーバ）を用いて増幅及び加工された音を出力する装置である。 FIG. 3 is a diagram showing an example of a data structure of selection information. FIG. 3 shows an example of selection information when the device is a hearing aid. Here, the hearing aid collects sound using, for example, a sound collecting unit (microphone), amplifies and processes the sound collected by using the processing unit, and amplifies and processes using the output unit (receiver). It is a device that outputs sound.
　図３の選択情報においては、「項目」には検査情報及び調査情報を識別する識別情報が記憶され、「質問内容」には検査情報及び調査情報を取得するために用いる質問内容を表す質問情報が記憶されている。 In the selection information of FIG. 3, the "item" stores the inspection information and the identification information that identifies the survey information, and the "question content" is the question information indicating the inspection information and the question content used to acquire the survey information. Is remembered.
　出力情報生成部２５は、推定部３が推定したパラメータデータの適合指標を、出力装置２２に出力するために用いる出力情報を生成する。そして、出力情報生成部２５は、生成した出力情報を出力装置２２へ出力する。 The output information generation unit 25 generates output information used to output the matching index of the parameter data estimated by the estimation unit 3 to the output device 22. Then, the output information generation unit 25 outputs the generated output information to the output device 22.
　図４は、パラメータデータのデータ構造の一例を示す図である。また、図４は、機器が補聴器の場合のパラメータデータを示している。図４の例では、パラメータデータは、例えば、補聴器に設けられている集音部、又は処理部、又は出力部、又はそれら各部の一つ以上の調整に用いるデータである。 FIG. 4 is a diagram showing an example of the data structure of the parameter data. Further, FIG. 4 shows parameter data when the device is a hearing aid. In the example of FIG. 4, the parameter data is, for example, data used for adjusting one or more of the sound collecting unit, the processing unit, the output unit, or each of these units provided in the hearing aid.
　また、パラメータデータが有するデータ（目的変数）それぞれは、図４に示すように、補聴器の種類、出力レベルごと（５０［ｄＢ］、７０［ｄＢ］、９０［ｄＢ］、最大出力レベル）の周波数特性値、騒音抑制強度、ハウリング抑制強度、指向性タイプ、衝撃音抑制度などが考えられる。 Further, as shown in FIG. 4, each of the data (objective variables) of the parameter data has a frequency of each type of hearing aid and output level (50 [dB], 70 [dB], 90 [dB], maximum output level). Characteristic values, noise suppression strength, howling suppression strength, directivity type, impact sound suppression degree, etc. can be considered.
　更に、図４の例では、パラメータデータは、パラメータデータが有する目的変数の項目を表す「項目」の情報と、推定したパラメータデータ（目的変数）を表す「パラメータ」の情報と、データの種別を表す「データ種別」の情報とが関連付けられている。 Further, in the example of FIG. 4, the parameter data includes "item" information representing the item of the objective variable possessed by the parameter data, "parameter" information representing the estimated parameter data (objective variable), and the type of data. It is associated with the information of the "data type" to be represented.
　パラメータデータの適合指標の分布は、例えば、図５に示すように表示することが考えられる。図５は、パラメータデータの適合指標の分布の表示例を示す図である。 The distribution of the matching index of the parameter data can be displayed as shown in FIG. 5, for example. FIG. 5 is a diagram showing a display example of the distribution of the conformity index of the parameter data.
　図４の「項目」、５０［ｄＢ］における周波数特性値の場合、目的変数は、周波数２５０［Ｈｚ］、５００［Ｈｚ］・・・それぞれに対応する、出力レベル３０［ｄＢ］、４０［ｄＢ］・・・となる。したがって、これら目的変数を決定する支援をするために、目的変数ごとに推定した確率分布を表示する。このように、表示する目的変数ごとに確率分布を表示することで、複数の目的変数の候補を、技能者に提示できる。 In the case of the frequency characteristic value in the "item" of FIG. 4, 50 [dB], the objective variable is the output level 30 [dB], 40 [dB] corresponding to the frequencies 250 [Hz], 500 [Hz], respectively. ] ... Therefore, in order to assist in determining these objective variables, the estimated probability distribution for each objective variable is displayed. By displaying the probability distribution for each objective variable to be displayed in this way, candidates for a plurality of objective variables can be presented to the technician.
　なお、図６に示すような表示をしてもよい。図６は、パラメータデータの適合指標の分布の表示例を示す図である。また、目的変数ごとに確率分布は、柱状図を用いて表示をしてもよい。更に、確率の大きさを、ヒートマップを用いて表示してもよい。 Note that the display as shown in FIG. 6 may be used. FIG. 6 is a diagram showing a display example of the distribution of the conformity index of the parameter data. Further, the probability distribution for each objective variable may be displayed using a column chart. Further, the magnitude of the probability may be displayed using a heat map.
　また、出力情報生成部２５は、選択部２４が選択した質問情報が表す質問内容を、出力装置２２に出力するために用いる出力情報を生成する。そして、出力情報生成部２５は、生成した出力情報を出力装置２２へ出力する。 Further, the output information generation unit 25 generates output information used to output the question content represented by the question information selected by the selection unit 24 to the output device 22. Then, the output information generation unit 25 outputs the generated output information to the output device 22.
　例えば、機器が補聴器の場合に、検査情報、調査情報が、図７、図８に示すような説明変数を有するものとする。図７は、検査情報のデータ構造の一例を示す図である。図８は、調査情報のデータ構造の一例を示す図である。 For example, when the device is a hearing aid, it is assumed that the examination information and the investigation information have explanatory variables as shown in FIGS. 7 and 8. FIG. 7 is a diagram showing an example of a data structure of inspection information. FIG. 8 is a diagram showing an example of the data structure of the survey information.
　図７の例では、機器が補聴器の場合、検査情報は、少なくとも気導オージオグラム、骨導オージオグラム、不快閾値、語音明瞭度のうち一つ以上の情報を有する。なお、図７の例では、検査情報には、聴力検査の項目を表す「項目」の情報と、聴力検査結果を表す「聴力検査結果」の情報と、データの種別を表す「データ種別」の情報とが関連付けられ、記憶装置に記憶されている。 In the example of FIG. 7, when the device is a hearing aid, the examination information has at least one or more information of air conduction audiogram, bone conduction audiogram, discomfort threshold, and speech intelligibility. In the example of FIG. 7, the test information includes "item" information indicating the item of the hearing test, "hearing test result" information indicating the hearing test result, and "data type" indicating the data type. The information is associated and stored in the storage device.
　図８の例では、調査情報は、例えば、少なくとも対象者の年齢、性別、職業、居所、家族構成、病歴、治療歴、機器の使用履歴、症状、身体的特徴に関する情報（例えば、身長、体重、耳音響など）のうちいずれか一つ以上の情報を有する。なお、耳音響は、耳の音響特性を表す情報である。 In the example of FIG. 8, the survey information is, for example, information on at least the subject's age, gender, occupation, whereabouts, family structure, medical history, treatment history, device usage history, symptoms, and physical characteristics (eg, height, weight). , Otoacoustic emission, etc.). The ear acoustics are information representing the acoustic characteristics of the ear.
　また、調査情報には、対象者の生活環境音を表す情報、趣向を表す情報、調整情報を有してもよい。生活環境音とは、対象者の普段の生活で聞こえている音を表す情報である。趣向は、対象者の音に対する趣向を表す情報である。調整情報は、機器の調整日時、調整場所、調整者識別番号などを表す情報である。更に、属性情報には、補聴器の種類を表す情報を含めてもよい。 In addition, the survey information may include information representing the living environment sound of the subject, information representing the taste, and adjustment information. Living environment sound is information representing the sound heard in the subject's daily life. The taste is information expressing the taste of the subject for the sound. The adjustment information is information indicating the adjustment date and time of the device, the adjustment place, the adjuster identification number, and the like. Further, the attribute information may include information indicating the type of hearing aid.
　なお、図８の例では、調整情報には、対象者に関係する調査（属性、背景など）の項目を表す「項目」の情報と、対象者の調査内容を表す「調査内容」の情報と、データの種別を表す「データ種別」の情報とが関連付けられ、記憶装置に記憶されている。 In the example of FIG. 8, the adjustment information includes "item" information representing the survey (attribute, background, etc.) items related to the target person, and "survey content" information representing the survey content of the target person. , The information of "data type" indicating the data type is associated with and stored in the storage device.
　このような説明変数を揃える場合に、例えば、不足している情報が、「語音明瞭度」「年齢」だとした場合、これらの説明変数に対応する、図４に示す質問内容「語音明瞭度の検査を結果の入力をしてください。」「年齢を入力してください。」を、出力装置２２に表示する。 When arranging such explanatory variables, for example, if the missing information is "speech intelligibility" and "age", the question content "speech intelligibility" shown in FIG. 4 corresponding to these explanatory variables. Please input the result of the inspection of. ”“ Enter the age. ”Is displayed on the output device 22.
　更に、出力情報生成部２５は、選択部２４が算出した情報量を、出力装置２２に出力するために用いる出力情報を生成する。そして、出力情報生成部２５は、生成した出力情報を出力装置２２へ出力する。 Further, the output information generation unit 25 generates output information used to output the amount of information calculated by the selection unit 24 to the output device 22. Then, the output information generation unit 25 outputs the generated output information to the output device 22.
　図９は、不足している情報に対応する情報量の表示例を示す図である。例えば、不足している情報に対する情報量を、図９に示すようなグラフを表示することが考えられる。 FIG. 9 is a diagram showing a display example of the amount of information corresponding to the missing information. For example, it is conceivable to display a graph as shown in FIG. 9 showing the amount of information for the missing information.
　次に、学習モデルの生成について説明する。
　図１０は、学習装置を有するシステムの一例を示す図である。図１０に示す学習装置３１は、機械学習を用いて学習モデル２６、２７、２８を生成する装置である。 Next, the generation of the learning model will be described.
FIG. 10 is a diagram showing an example of a system having a learning device. The learning device 31 shown in FIG. 10 is a device that generates learning models 26, 27, and 28 using machine learning.
　図１０に示す学習装置３１を有するシステムは、学習装置３１に加えて、記憶装置３２を有する。また、学習装置３１は、取得部３３と、分類部３４と、分類部３５と、生成部３６とを有する。 The system having the learning device 31 shown in FIG. 10 has a storage device 32 in addition to the learning device 31. Further, the learning device 31 has an acquisition unit 33, a classification unit 34, a classification unit 35, and a generation unit 36.
　記憶装置３２は、過去において取得した説明変数（検査情報、調査情報）と、目的変数（パラメータデータ）とを記憶している。 The storage device 32 stores explanatory variables (inspection information, survey information) acquired in the past and objective variables (parameter data).
　過去において取得した複数の検査情報とは、過去において、複数の機器の利用者が、利用者ごとに実施した検査の結果を表す情報である。過去において取得した複数の調整情報とは、過去において、複数の機器の利用者が、利用者ごとに取得した調整情報である。 The plurality of inspection information acquired in the past is information representing the results of inspections conducted by users of a plurality of devices in the past for each user. The plurality of adjustment information acquired in the past is the adjustment information acquired by the users of a plurality of devices for each user in the past.
　過去において取得したパラメータデータとは、過去において、複数の機器の利用者に対して、技能レベルの高い技能者が実施したフィッティングにおいて、機器の調整に用いたパラメータデータである。また、過去において取得したパラメータデータは、過去において、複数の機器の利用者に対して、フィッティング支援装置１を用いて実施したフィッティングにおいて、機器の調整に用いたパラメータデータである。 The parameter data acquired in the past is the parameter data used for adjusting the equipment in the fitting performed by the technician with a high skill level for the users of a plurality of equipment in the past. Further, the parameter data acquired in the past is the parameter data used for adjusting the equipment in the fitting performed by using the fitting support device 1 for the users of a plurality of equipment in the past.
　学習装置について説明する。
　取得部３３は、過去において取得した学習データを取得する。具体的には、取得部３３は、記憶装置３２から、過去において取得した複数の検査情報、調整情報、パラメータデータなどの学習データを取得し、取得した学習データを分類部３４へ送信する。 The learning device will be described.
The acquisition unit 33 acquires the learning data acquired in the past. Specifically, the acquisition unit 33 acquires learning data such as a plurality of inspection information, adjustment information, and parameter data acquired in the past from the storage device 32, and transmits the acquired learning data to the classification unit 34.
　分類部３４は、受信した学習データを分類する。具体的には、分類部３４は、まず、過去において、利用者がフィッティングに対して満足したか否かを表すスコア（満足度）と、あらかじめ設定した閾値とを比較する。続いて、分類部３４は、スコアが閾値以上である場合、閾値以上のスコアに関連付けられている学習データを分類する。 The classification unit 34 classifies the received learning data. Specifically, the classification unit 34 first compares a score (satisfaction level) indicating whether or not the user is satisfied with the fitting in the past with a preset threshold value. Subsequently, when the score is equal to or higher than the threshold value, the classification unit 34 classifies the learning data associated with the score equal to or higher than the threshold value.
　このように、分類部３４は、満足度が高い学習データのみを用いて学習モデル２６、２７、２８に学習をさせることができるため、学習モデル２６、２７、２８を用いて推定したパラメータデータは、利用者の満足度が高くなる。 In this way, the classification unit 34 can make the learning models 26, 27, 28 train using only the learning data having a high degree of satisfaction, so that the parameter data estimated using the learning models 26, 27, 28 can be used. , User satisfaction is high.
　分類部３５は、分類部３４で分類した学習データを更に分類する。具体的には、分類部３５は、分類部３４で分類した学習データに対して、更にクラスタリング処理を実行する。 The classification unit 35 further classifies the learning data classified by the classification unit 34. Specifically, the classification unit 35 further executes a clustering process on the learning data classified by the classification unit 34.
　クラスタリング処理では、高次元の情報を低次元の情報へ変換する。その理由は、文章や時系列信号で表現されうる説明変数は膨大な数値・カテゴリ次元を持つ変数となってしまい、そのまま学習をすると、計算資源量が増加する。そこで、文字列データに対しては各単語の有無の評価、ＴＦ－ＩＤＦなどを適用して、数値ベクトルへ変換し、このベクトル化したデータに対して、例えば、ｋ－平均法、深層学習などを用いて高次元の情報を低次元の情報へ変換をすることが考えられる。 In the clustering process, high-dimensional information is converted into low-dimensional information. The reason is that explanatory variables that can be expressed by sentences and time-series signals become variables with enormous numerical and categorical dimensions, and if learning is performed as it is, the amount of computational resources will increase. Therefore, evaluation of the presence or absence of each word, TF-IDF, etc. are applied to the character string data to convert it into a numerical vector, and for this vectorized data, for example, k-means clustering, deep learning, etc. It is conceivable to convert high-dimensional information into low-dimensional information using.
　一例として、まず、高次元の情報（例えば、文字列、時系列信号など）を、深層ニューラルネットワークに入力して、ベクトル値に変換する。続いて、変換したベクトル値を、ｋ近傍法を用いて、低次元の情報（例えば、ラベルなど）に変換する。 As an example, first, high-dimensional information (for example, character string, time series signal, etc.) is input to a deep neural network and converted into a vector value. Subsequently, the converted vector value is converted into low-dimensional information (for example, a label) by using the k-nearest neighbor method.
　なお、ドメイン知識から作成されたラベルを用いて教師あり学習で分類を行う他にも、ラベルが事前に付与されていない場合でも、ｋ－平均法などによるクラスタリング手法を用いることで教師無し学習でラベルを作ることができる。 In addition to classifying by supervised learning using labels created from domain knowledge, even if labels are not given in advance, unsupervised learning can be performed by using a clustering method such as the k-means method. You can make labels.
　このように、分類部３４においては、満足度が高い学習データのみを用いて学習モデル２６、２７、２８に学習をさせることができるため、学習モデル２６、２７、２８を用いて推定したパラメータデータは、利用者の満足度が高くなる。更に、学習データを削減することができるので、学習にかかる時間の短縮、メモリ使用量の低減など、計算資源量を削減することができる。 As described above, in the classification unit 34, since the learning models 26, 27, and 28 can be trained using only the learning data having a high degree of satisfaction, the parameter data estimated using the learning models 26, 27, 28. Will increase user satisfaction. Further, since the learning data can be reduced, the amount of computational resources can be reduced, such as shortening the learning time and reducing the memory usage.
　また、分類部３５においては、分類部３４で分類した学習データに対してクラスタリング処理を実行することで、学習する場合において、更に計算資源量を削減することができる。 Further, in the classification unit 35, by executing the clustering process on the learning data classified by the classification unit 34, the amount of computational resources can be further reduced in the case of learning.
　生成部３６は、分類部３５で分類した学習データを用いて機械学習をさせて、学習モデル２６、２７、２８を生成し、生成した学習モデル２６、２７、２８を推定部３に記憶する。又は、生成部３６は、学習モデル２６、２７、２８を、学習装置３１を有するシステム、又はフィッティング支援装置１を有するシステム、又はそれら以外の記憶装置に記憶してもよい。 The generation unit 36 performs machine learning using the learning data classified by the classification unit 35 to generate learning models 26, 27, 28, and stores the generated learning models 26, 27, 28 in the estimation unit 3. Alternatively, the generation unit 36 may store the learning models 26, 27, and 28 in a system having the learning device 31, a system having the fitting support device 1, or a storage device other than the system.
　なお、機械学習は、例えば、教師あり学習、半教師あり学習などが考えられる。例えば、回帰（最小二乗法、ベイズ線型回帰、ランダムフォレストなどの回帰手法）、多クラス分類（決定木などのアルゴリズム）などの学習を用いることが考えられる。なお、上述した機械学習に限定されるものではなく、上述した機器学習以外を用いてもよい。 Note that machine learning can be, for example, supervised learning or semi-supervised learning. For example, it is conceivable to use learning such as regression (least squares method, Bayesian linear regression, regression method such as random forest), and multiclass classification (algorithm such as decision tree). It should be noted that the present invention is not limited to the above-mentioned machine learning, and other than the above-mentioned device learning may be used.
　ただし、生成部３６の入力は、分類部３５で分類された学習データを用いてもよいし、又は分類部３４で分類された学習データを用いてもよいし、又は分類されていない学習データを用いてもよい。 However, as the input of the generation unit 36, the learning data classified by the classification unit 35 may be used, the learning data classified by the classification unit 34 may be used, or the unclassified learning data may be used. You may use it.
　また、学習モデル２６、２７、２８を生成するために用いる学習データは、過去に適合状態と判断された検査情報と、調整情報と、パラメータデータとを用いてもよい。 Further, as the learning data used to generate the learning models 26, 27, 28, inspection information, adjustment information, and parameter data that have been determined to be in conformity in the past may be used.
　このように、学習装置３１は、過去において取得した検査情報だけでなく、過去において取得した調整情報、パラメータデータなども加味した、学習モデル２６、２７、２８を生成できる。特に、技能レベルの高い技能者が実施したフィッティングの結果を入力として、学習モデル２６、２７、２８を生成できる。 In this way, the learning device 31 can generate learning models 26, 27, and 28 that take into account not only the inspection information acquired in the past but also the adjustment information and parameter data acquired in the past. In particular, learning models 26, 27, and 28 can be generated by inputting the results of fitting performed by a technician with a high skill level.
［装置動作］
　次に、本発明の実施の形態におけるフィッティング支援装置の動作について図１１を用いて説明する。図１１は、フィッティング支援装置の動作の一例を示す図である。以下の説明においては、適宜図２から９を参照する。また、本実施の形態では、フィッティング支援装置を動作させることによって、フィッティング支援方法が実施される。よって、本実施の形態におけるフィッティング支援方法の説明は、以下のフィッティング支援装置の動作説明に代える。 [Device operation]
Next, the operation of the fitting support device according to the embodiment of the present invention will be described with reference to FIG. FIG. 11 is a diagram showing an example of the operation of the fitting support device. In the following description, FIGS. 2 to 9 will be referred to as appropriate. Further, in the present embodiment, the fitting support method is implemented by operating the fitting support device. Therefore, the description of the fitting support method in the present embodiment will be replaced with the following description of the operation of the fitting support device.
　また、本発明の実施の形態における学習装置の動作について図１２を用いて説明する。図１２は、学習装置の動作の一例を示す図である。以下の説明においては、適宜図１０を参照する。また、本実施の形態では、学習装置を動作させることによって、学習方法が実施される。よって、本実施の形態における学習方法の説明は、以下の学習装置の動作説明に代える。 Further, the operation of the learning device according to the embodiment of the present invention will be described with reference to FIG. FIG. 12 is a diagram showing an example of the operation of the learning device. In the following description, FIG. 10 will be referred to as appropriate. Further, in the present embodiment, the learning method is implemented by operating the learning device. Therefore, the description of the learning method in the present embodiment is replaced with the following description of the operation of the learning device.
　フィッティング支援装置の動作について説明する。
　図１１に示すように、最初に、取得部２は、運用フェーズにおいて、入力装置２１から対象者の検査情報、調査情報（説明変数）を取得する（ステップＡ１）。具体的には、ステップＡ１において、取得部２は、入力装置２１から送信された、対象者に対して実施した検査の結果を表す検査情報と、対象者に関係する調査の結果を表す調査情報とを受信する。 The operation of the fitting support device will be described.
As shown in FIG. 11, first, the acquisition unit 2 acquires the inspection information and the investigation information (explanatory variable) of the target person from the input device 21 in the operation phase (step A1). Specifically, in step A1, the acquisition unit 2 receives the inspection information transmitted from the input device 21 indicating the result of the inspection performed on the target person and the survey information representing the result of the investigation related to the target person. And receive.
　次に、推定部３は、運用フェーズにおいて、取得した検査情報及び調査情報を入力として、パラメータデータの適合度を表す適合指標及びその分布を推定する（ステップＡ２）。具体的には、ステップＡ２において、推定部３は、まず、検査情報及び調査情報（説明変数）を取得する。続いて、ステップＡ２において、推定部３は、取得した検査情報及び調査情報（説明変数）を、学習モデル２６に入力し、パラメータデータの適合度を表す適合指標及びその分布を推定する。 Next, in the operation phase, the estimation unit 3 uses the acquired inspection information and survey information as inputs to estimate the goodness-of-fit index representing the goodness of fit of the parameter data and its distribution (step A2). Specifically, in step A2, the estimation unit 3 first acquires inspection information and survey information (explanatory variables). Subsequently, in step A2, the estimation unit 3 inputs the acquired inspection information and survey information (explanatory variable) into the learning model 26, and estimates the goodness-of-fit index representing the goodness of fit of the parameter data and its distribution.
　次に、選択部２４は、運用フェーズにおいて、フィッティングに用いる検査情報、又は調査情報、又はそれら両方に不足している情報がある場合、不足している情報を補うために用いる質問情報を選択する（ステップＡ３）。具体的には、ステップＡ３において、選択部２４は、まず、フィッティングに用いる検査情報、又は調査情報、又はそれら両方（説明変数）に、不足している情報があるか否かを検出する。 Next, in the operation phase, the selection unit 24 selects question information to be used to supplement the missing information when there is missing information in the inspection information, the survey information, or both of them used for fitting. (Step A3). Specifically, in step A3, the selection unit 24 first detects whether or not there is insufficient information in the inspection information used for fitting, the survey information, or both (explanatory variables).
　続いて、ステップＡ３において、選択部２４は、不足している情報（説明変数）がある場合、不足している情報を得るために用いる情報（例えば、説明変数ごとに情報取得済みか否かを表すバイナリ値を要素とするベクトルなど）を、学習モデル２７に入力し、不足している情報の確率分布を推定する。また、ステップＡ３において、選択部２４は、不足している情報がある場合、不足している情報を得るために用いる情報（例えば、説明変数ごとに情報取得済みか否かを表すバイナリ値を要素とするベクトルなど）を、学習モデル２８に入力し、パラメータデータと不足している情報との同時分布を推定する。 Subsequently, in step A3, when there is missing information (explanatory variable), the selection unit 24 determines whether or not the information used to obtain the missing information (for example, information has been acquired for each explanatory variable). (A vector having the represented binary value as an element, etc.) is input to the learning model 27, and the probability distribution of the missing information is estimated. Further, in step A3, when there is missing information, the selection unit 24 elements the information used to obtain the missing information (for example, a binary value indicating whether or not the information has been acquired for each explanatory variable). (Such as the vector to be used) is input to the learning model 28, and the simultaneous distribution of the parameter data and the missing information is estimated.
　続いて、ステップＡ３において、選択部２４は、不足している情報ごとに、推定した確率分布と、同時分布とを用いて、情報量（重要度）を算出する。続いて、ステップＡ３において、選択部２４は、不足している情報ごと、パラメータデータ（目的変数）との相関を表す情報量を算出し、この情報量に基づいて、不足している情報の中から一つ以上の情報を選択する。 Subsequently, in step A3, the selection unit 24 calculates the amount of information (importance) for each missing information by using the estimated probability distribution and the joint distribution. Subsequently, in step A3, the selection unit 24 calculates the amount of information representing the correlation with the parameter data (objective variable) for each missing information, and based on this amount of information, among the missing information. Select one or more pieces of information from.
　その後、ステップＡ３において、選択部２４は、選択した不足している情報に基づいて、説明変数を識別する識別情報と質問情報とが関連付けられた選択情報を参照し、質問情報を選択する。 After that, in step A3, the selection unit 24 refers to the selection information associated with the identification information for identifying the explanatory variable and the question information based on the selected missing information, and selects the question information.
　次に、出力情報生成部２５は、推定部３が推定したパラメータデータの適合指標を、出力装置２２に出力するために用いる出力情報を生成する（ステップＡ４）。そして、出力情報生成部２５は、生成した出力情報を出力装置２２へ出力する（ステップＡ５）。 Next, the output information generation unit 25 generates output information used to output the matching index of the parameter data estimated by the estimation unit 3 to the output device 22 (step A4). Then, the output information generation unit 25 outputs the generated output information to the output device 22 (step A5).
　なお、パラメータデータの適合指標は、例えば、図５、図６に示すように表示することが考えられる。また、目的変数ごとに確率分布は、柱状図を用いて表示をしてもよい。更に、確率の大きさを、ヒートマップ又は等高線図などを用いて表示してもよい。 It is conceivable that the matching index of the parameter data is displayed as shown in FIGS. 5 and 6, for example. Further, the probability distribution for each objective variable may be displayed using a column chart. Further, the magnitude of the probability may be displayed by using a heat map, a contour map, or the like.
　また、ステップＡ４において、出力情報生成部２５は、選択部２４が選択した質問情報が表す質問内容を、出力装置２２に出力するために用いる出力情報を生成する。そして、ステップＡ５において、出力情報生成部２５は、生成した出力情報を出力装置２２へ出力する。 Further, in step A4, the output information generation unit 25 generates output information used to output the question content represented by the question information selected by the selection unit 24 to the output device 22. Then, in step A5, the output information generation unit 25 outputs the generated output information to the output device 22.
　更に、ステップＡ４において、出力情報生成部２５は、選択部２４が算出した情報量を、出力装置２２に出力するために用いる出力情報を生成する。そして、ステップＡ５において、出力情報生成部２５は、生成した出力情報を出力装置２２へ出力する。 Further, in step A4, the output information generation unit 25 generates output information used to output the amount of information calculated by the selection unit 24 to the output device 22. Then, in step A5, the output information generation unit 25 outputs the generated output information to the output device 22.
　このように、パラメータデータを決定するために必要な検査情報及び調査情報（説明変数）が揃うまで、ステップＡ１からＡ５の処理を繰り返す。 In this way, the processes of steps A1 to A5 are repeated until the inspection information and the survey information (explanatory variables) necessary for determining the parameter data are prepared.
　なお、不足している情報があった場合には、ステップＡ１において、質問内容を用いて取得した不足している情報を追加して、再度、ステップＡ２からＡ５の処理を実行する。　　　 If there is missing information, in step A1, the missing information acquired by using the question content is added, and the processes of steps A2 to A5 are executed again.
　学習装置の動作について説明する。
　図１２に示すように、最初に、取得部３３は、過去において取得した学習データを取得する（ステップＢ１）。具体的には、ステップＢ１において、取得部３３は、記憶装置３２から、過去において取得した複数の検査情報、調査情報、パラメータデータなどの学習データを取得し、取得した学習データを分類部３４へ送信する。 The operation of the learning device will be described.
As shown in FIG. 12, first, the acquisition unit 33 acquires the learning data acquired in the past (step B1). Specifically, in step B1, the acquisition unit 33 acquires learning data such as a plurality of inspection information, survey information, and parameter data acquired in the past from the storage device 32, and transfers the acquired learning data to the classification unit 34. Send.
　次に、分類部３４は、受信した学習データを分類する（ステップＢ２）。具体的には、ステップＢ２において、分類部３４は、まず、過去において、利用者がフィッティングに対して満足したか否かを表すスコア（満足度）と、あらかじめ設定した閾値とを比較する。続いて、ステップＢ２において、分類部３４は、スコアが閾値以上である場合、閾値以上のスコアに関連付けられている学習データを分類する。 Next, the classification unit 34 classifies the received learning data (step B2). Specifically, in step B2, the classification unit 34 first compares a score (satisfaction level) indicating whether or not the user is satisfied with the fitting in the past with a preset threshold value. Subsequently, in step B2, when the score is equal to or higher than the threshold value, the classification unit 34 classifies the learning data associated with the score equal to or higher than the threshold value.
　次に、分類部３５は、ステップＢ２において分類した学習データを更に分類する（ステップＢ３）。具体的には、ステップＢ３において、分類部３５は、ステップＢ２において分類した学習データに対して、更にクラスタリング処理を実行する。クラスタリング処理では、高次元の情報を低次元の情報へ変換する。 Next, the classification unit 35 further classifies the learning data classified in step B2 (step B3). Specifically, in step B3, the classification unit 35 further executes a clustering process on the learning data classified in step B2. In the clustering process, high-dimensional information is converted into low-dimensional information.
　次に、生成部３６は、分類部３５で分類した学習データを用いて機械学習をさせて、学習モデル２６、２７、２８を生成し（ステップＢ４）、生成した学習モデル２６、２７、２８を推定部３に記憶する（ステップＢ５）。又は、ステップＢ５において、生成部３６は、学習モデル２６、２７、２８を、学習装置３１を有するシステム、又はフィッティング支援装置１を有するシステム、又はそれら以外の記憶装置に記憶してもよい。 Next, the generation unit 36 causes machine learning using the learning data classified by the classification unit 35 to generate learning models 26, 27, 28 (step B4), and generates the generated learning models 26, 27, 28. It is stored in the estimation unit 3 (step B5). Alternatively, in step B5, the generation unit 36 may store the learning models 26, 27, and 28 in a system having the learning device 31, a system having the fitting support device 1, or a storage device other than those.
　ただし、生成部３６の入力は、分類部３５で分類された学習データを用いてもよいし、又は分類部３４で分類された学習データを用いてもよいし、又は分類されていない学習データを用いてもよい。 However, as the input of the generation unit 36, the learning data classified by the classification unit 35 may be used, the learning data classified by the classification unit 34 may be used, or the unclassified learning data may be used. You may use it.
　また、学習モデル２６、２７、２８を生成するために用いる学習データは、過去に適合状態と判断された検査情報と、調整情報と、パラメータデータとを用いてもよい。 Further, as the learning data used to generate the learning models 26, 27, 28, inspection information, adjustment information, and parameter data that have been determined to be in conformity in the past may be used.
［本実施の形態の効果］
　以上のように本実施の形態によれば、フィッティング支援装置１は、パラメータデータの適合指標及びその分布を推定できるので、推定したパラメータデータの適合指標及びその分布を技能者に提示できる。そのため、技能者に複数の候補となるパラメータを提示できる。 [Effect of this embodiment]
As described above, according to the present embodiment, the fitting support device 1 can estimate the matching index of the parameter data and its distribution, so that the matching index of the estimated parameter data and its distribution can be presented to the technician. Therefore, it is possible to present a plurality of candidate parameters to the technician.
　具体的には、フィッティング支援装置１は、パラメータデータの適合指標の分布として確率分布を出力装置に出力することで、技能レベルの低い技能者でも、出力された確率分布を参考にして、対象者に対して適切なフィッティングを実施することができる。そのため、技能レベルの低い技能者の技能を向上させることができるので、フィッティングの精度を向上させることができる。 Specifically, the fitting support device 1 outputs a probability distribution to the output device as the distribution of the matching index of the parameter data, so that even a technician with a low skill level can refer to the output probability distribution and the target person. Appropriate fitting can be performed for. Therefore, the skill of a technician with a low skill level can be improved, and the accuracy of fitting can be improved.
　また、パラメータデータの適合指標及びその分布を推定し、複数の候補となるパラメータを技能者に提示することで、具体的なフィッティングの方針を提示できるため、技能者がフィッティングのために要する時間を短縮できる。また、候補となるパラメータを用いることで、パラメータを対象者に適合し易くできる。 In addition, by estimating the matching index of the parameter data and its distribution and presenting a plurality of candidate parameters to the technician, a specific fitting policy can be presented, so that the technician can take the time required for fitting. Can be shortened. Further, by using the candidate parameters, the parameters can be easily adapted to the target person.
　また、調査内容を提示できるので、具体的なフィッティングの方針が明確にできる。そのため、技能者がフィッティングのために要する時間を短縮できる。また、必要な説明変数をできるだけ揃えることができるので、対象者に適合したパラメータを推定する精度を向上させることができる。 Also, since the survey contents can be presented, the specific fitting policy can be clarified. Therefore, the time required for the technician to perform the fitting can be shortened. In addition, since the necessary explanatory variables can be arranged as much as possible, the accuracy of estimating the parameters suitable for the target person can be improved.
［プログラム］
　本発明の実施の形態におけるパラメータデータの適合指標及びその分布を推定するためのプログラムは、コンピュータに、図１１に示すステップＡ１からＡ５を実行させるプログラムであればよい。このプログラムをコンピュータにインストールし、実行することによって、本実施の形態におけるフィッティング支援装置とフィッティング支援方法とを実現することができる。この場合、コンピュータのプロセッサは、取得部２、推定部３、選択部２４、出力情報生成部２５として機能し、処理を行なう。 [program]
The program for estimating the matching index of the parameter data and its distribution in the embodiment of the present invention may be a program that causes a computer to execute steps A1 to A5 shown in FIG. By installing this program on a computer and executing it, the fitting support device and the fitting support method according to the present embodiment can be realized. In this case, the computer processor functions as an acquisition unit 2, an estimation unit 3, a selection unit 24, and an output information generation unit 25, and performs processing.
　また、本実施の形態におけるパラメータデータの適合指標及びその分布を推定するためのプログラムは、複数のコンピュータによって構築されたコンピュータシステムによって実行されてもよい。この場合は、例えば、各コンピュータが、それぞれ、取得部２、推定部３、選択部２４、出力情報生成部２５のいずれかとして機能してもよい。 Further, the program for estimating the matching index of the parameter data and its distribution in the present embodiment may be executed by a computer system constructed by a plurality of computers. In this case, for example, each computer may function as any of the acquisition unit 2, the estimation unit 3, the selection unit 24, and the output information generation unit 25, respectively.
　更に、本発明の実施の形態における学習モデルを生成するためのプログラムは、コンピュータに、図１２に示すステップＢ１からＢ５を実行させるプログラムであればよい。このプログラムをコンピュータにインストールし、実行することによって、本実施の形態における学習装置と学習方法とを実現することができる。この場合、コンピュータのプロセッサは、取得部３３、分類部３４、分類部３５、生成部３６として機能し、処理を行なう。 Further, the program for generating the learning model according to the embodiment of the present invention may be a program that causes a computer to execute steps B1 to B5 shown in FIG. By installing this program on a computer and executing it, the learning device and the learning method according to the present embodiment can be realized. In this case, the computer processor functions as an acquisition unit 33, a classification unit 34, a classification unit 35, and a generation unit 36, and performs processing.
　また、本実施の形態における学習モデルを生成するためのプログラムは、複数のコンピュータによって構築されたコンピュータシステムによって実行されてもよい。この場合は、例えば、各コンピュータが、それぞれ、取得部３３、分類部３４、分類部３５、生成部３６のいずれかとして機能してもよい。 Further, the program for generating the learning model in the present embodiment may be executed by a computer system constructed by a plurality of computers. In this case, for example, each computer may function as any of the acquisition unit 33, the classification unit 34, the classification unit 35, and the generation unit 36, respectively.
［物理構成］
　ここで、実施の形態におけるプログラムを実行することによって、フィッティング支援装置又は学習装置を実現するコンピュータについて図１３を用いて説明する。図１３は、本発明の実施の形態におけるフィッティング支援装置又は学習装置を実現するコンピュータの一例を示すブロック図である。 [Physical configuration]
Here, a computer that realizes a fitting support device or a learning device by executing the program in the embodiment will be described with reference to FIG. FIG. 13 is a block diagram showing an example of a computer that realizes the fitting support device or the learning device according to the embodiment of the present invention.
　図１３に示すように、コンピュータ１１０は、ＣＰＵ（Central Processing Unit）１１１と、メインメモリ１１２と、記憶装置１１３と、入力インターフェイス１１４と、表示コントローラ１１５と、データリーダ／ライタ１１６と、通信インターフェイス１１７とを備える。これらの各部は、バス１２１を介して、互いにデータ通信可能に接続される。なお、コンピュータ１１０は、ＣＰＵ１１１に加えて、又はＣＰＵ１１１に代えて、ＧＰＵ（Graphics Processing Unit）、又はＦＰＧＡ（Field-Programmable Gate Array）を備えていてもよい。 As shown in FIG. 13, the computer 110 includes a CPU (Central Processing Unit) 111, a main memory 112, a storage device 113, an input interface 114, a display controller 115, a data reader / writer 116, and a communication interface 117. And. Each of these parts is connected to each other via a bus 121 so as to be capable of data communication. The computer 110 may include a GPU (Graphics Processing Unit) or an FPGA (Field-Programmable Gate Array) in addition to the CPU 111 or in place of the CPU 111.
　ＣＰＵ１１１は、記憶装置１１３に格納された、本実施の形態におけるプログラム（コード）をメインメモリ１１２に展開し、これらを所定順序で実行することにより、各種の演算を実施する。メインメモリ１１２は、典型的には、ＤＲＡＭ（Dynamic Random Access Memory）などの揮発性の記憶装置である。また、本実施の形態におけるプログラムは、コンピュータ読み取り可能な記録媒体１２０に格納された状態で提供される。なお、本実施の形態におけるプログラムは、通信インターフェイス１１７を介して接続されたインターネット上で流通するものであってもよい。 The CPU 111 expands the programs (codes) of the present embodiment stored in the storage device 113 into the main memory 112 and executes them in a predetermined order to perform various operations. The main memory 112 is typically a volatile storage device such as a DRAM (Dynamic Random Access Memory). Further, the program according to the present embodiment is provided in a state of being stored in a computer-readable recording medium 120. The program in the present embodiment may be distributed on the Internet connected via the communication interface 117.
　また、記憶装置１１３の具体例としては、ハードディスクドライブの他、フラッシュメモリなどの半導体記憶装置があげられる。入力インターフェイス１１４は、ＣＰＵ１１１と、キーボード及びマウスといった入力機器１１８との間のデータ伝送を仲介する。表示コントローラ１１５は、ディスプレイ装置１１９と接続され、ディスプレイ装置１１９での表示を制御する。 Further, specific examples of the storage device 113 include a semiconductor storage device such as a flash memory in addition to a hard disk drive. The input interface 114 mediates data transmission between the CPU 111 and an input device 118 such as a keyboard and mouse. The display controller 115 is connected to the display device 119 and controls the display on the display device 119.
　データリーダ／ライタ１１６は、ＣＰＵ１１１と記録媒体１２０との間のデータ伝送を仲介し、記録媒体１２０からのプログラムの読み出し、及びコンピュータ１１０における処理結果の記録媒体１２０への書き込みを実行する。通信インターフェイス１１７は、ＣＰＵ１１１と、他のコンピュータとの間のデータ伝送を仲介する。 The data reader / writer 116 mediates the data transmission between the CPU 111 and the recording medium 120, reads the program from the recording medium 120, and writes the processing result in the computer 110 to the recording medium 120. The communication interface 117 mediates data transmission between the CPU 111 and another computer.
　また、記録媒体１２０の具体例としては、ＣＦ（Compact Flash（登録商標））及びＳＤ（Secure Digital）などの汎用的な半導体記憶デバイス、フレキシブルディスク（Flexible Disk）などの磁気記録媒体、又はＣＤ－ＲＯＭ（Compact Disk Read Only Memory）などの光学記録媒体があげられる。 Specific examples of the recording medium 120 include a general-purpose semiconductor storage device such as CF (CompactFlash (registered trademark)) and SD (SecureDigital), a magnetic recording medium such as a flexible disk, or a CD-. Examples include optical recording media such as ROM (CompactDiskReadOnlyMemory).
　なお、本実施の形態におけるフィッティング支援装置１又は学習装置３１は、プログラムがインストールされたコンピュータではなく、各部に対応したハードウェアを用いることによっても実現可能である。更に、フィッティング支援装置１又は学習装置３１は、一部がプログラムで実現され、残りの部分がハードウェアで実現されていてもよい。 Note that the fitting support device 1 or the learning device 31 in the present embodiment can also be realized by using hardware corresponding to each part instead of the computer on which the program is installed. Further, the fitting support device 1 or the learning device 31 may be partially realized by a program and the rest may be realized by hardware.
［付記］
　以上の実施の形態に関し、更に以下の付記を開示する。上述した実施の形態の一部又は全部は、以下に記載する（付記１）から（付記１８）により表現することができるが、以下の記載に限定されるものではない。 [Additional Notes]
The following additional notes will be further disclosed with respect to the above embodiments. A part or all of the above-described embodiments can be expressed by the following descriptions (Appendix 1) to (Appendix 18), but are not limited to the following descriptions.
（付記１）
　対象者に対して実施した検査の結果を表す検査情報と、前記対象者に関係する調査の結果を表す調査情報とを取得する、取得部と、
　前記対象者に機器を適合させるために用いるパラメータデータのフィッティングにおいて、取得した前記検査情報と前記調査情報とを入力して、前記パラメータデータの適合度を表す適合指標及びその分布を推定する、推定部と、
　を有することを特徴とするフィッティング支援装置。 (Appendix 1)
An acquisition unit that acquires inspection information representing the results of inspections conducted on the target person and survey information representing the results of the survey related to the target person.
In the fitting of the parameter data used to fit the device to the target person, the acquired inspection information and the survey information are input to estimate the goodness of fit index representing the goodness of fit of the parameter data and its distribution. Department and
A fitting support device characterized by having.
（付記２）
　付記１に記載のフィッティング支援装置であって、
　前記パラメータデータの適合指標及びその分布を表す出力情報を生成して出力装置に出力する、出力情報生成部と、
　を有することを特徴とするフィッティング支援装置。 (Appendix 2)
The fitting support device according to Appendix 1.
An output information generator that generates output information representing the matching index of the parameter data and its distribution and outputs it to the output device.
A fitting support device characterized by having.
（付記３）
　付記２に記載のフィッティング支援装置であって、
　前記推定部は、過去に取得した複数の利用者の検査情報と、過去に取得した複数の利用者の調査情報と、過去に前記機器に設定されたパラメータデータとを入力として、機械学習により生成された、前記パラメータデータの適合指標及びその分布の推定に用いる学習モデルを有する
　ことを特徴とするフィッティング支援装置。 (Appendix 3)
The fitting support device described in Appendix 2
The estimation unit is generated by machine learning by inputting inspection information of a plurality of users acquired in the past, survey information of a plurality of users acquired in the past, and parameter data set in the device in the past. A fitting support device having a learning model used for estimating a matching index of the parameter data and its distribution.
（付記４）
　付記３に記載のフィッティング支援装置であって、
　フィッティングに用いる前記検査情報、又は前記調査情報、又はそれら両方に不足している情報がある場合、前記不足している情報を補うために、前記不足している情報に関連付けられた質問情報を選択する、選択部と、
　前記出力情報生成部は、前記質問情報を用いて質問内容を表す前記出力情報を生成して出力装置に出力する
　ことを特徴とするフィッティング支援装置。 (Appendix 4)
The fitting support device described in Appendix 3
If the inspection information used for fitting, the survey information, or both of them have missing information, the question information associated with the missing information is selected to make up for the missing information. To do, the selection part,
The output information generation unit is a fitting support device characterized in that the output information representing the question content is generated using the question information and output to the output device.
（付記５）
　付記４に記載のフィッティング支援装置であって、
　前記選択部は、前記不足している情報と前記パラメータデータとの相関を表す情報量を、前記不足している情報に対して算出し、算出した前記情報量に基づいて、前記不足している情報を選択する
　ことを特徴とするフィッティング支援装置。 (Appendix 5)
The fitting support device according to Appendix 4.
The selection unit calculates an amount of information representing the correlation between the deficient information and the parameter data with respect to the deficient information, and based on the calculated amount of information, the deficiency A fitting support device characterized by selecting information.
（付記６）
　付記４又は５に記載のフィッティング支援装置であって、
　前記推定部は、前記質問情報を用いて前記不足している情報が取得された場合、取得された前記不足している情報を追加して、再度、前記パラメータデータの適合指標及びその分布を推定する
　を有することを特徴とするフィッティング支援装置。 (Appendix 6)
The fitting support device according to Appendix 4 or 5.
When the missing information is acquired using the question information, the estimation unit adds the acquired missing information and estimates the matching index of the parameter data and its distribution again. A fitting support device characterized by having a device.
（付記７）
（ａ）対象者に対して実施した検査の結果を表す検査情報と、前記対象者に関係する調査の結果を表す調査情報とを取得する、ステップと、
（ｂ）前記対象者に機器を適合させるために用いるパラメータデータのフィッティングにおいて、取得した前記検査情報と前記調査情報とを入力して、前記パラメータデータの適合度を表す適合指標及びその分布を推定する、ステップと、
　を有することを特徴とするフィッティング支援方法。 (Appendix 7)
(A) A step of acquiring inspection information representing the results of an inspection conducted on a subject and survey information representing the results of a survey related to the subject.
(B) In fitting the parameter data used to fit the device to the target person, the acquired inspection information and the survey information are input to estimate the goodness of fit index representing the goodness of fit of the parameter data and its distribution. To do, step and
A fitting support method characterized by having.
（付記８）
　付記７に記載のフィッティング支援方法であって、
（ｃ）前記パラメータデータの適合指標及びその分布を表す出力情報を生成して出力装置に出力する、ステップと、
　を有することを特徴とするフィッティング支援方法。 (Appendix 8)
The fitting support method described in Appendix 7
(C) A step of generating output information representing the conformity index of the parameter data and its distribution and outputting it to the output device.
A fitting support method characterized by having.
（付記９）
　付記８に記載のフィッティング支援方法であって、
　前記（ｂ）のステップにおいて、過去に取得した複数の利用者の検査情報と、過去に取得した複数の利用者の調査情報と、過去に前記機器に設定されたパラメータデータとを入力として、機械学習により生成された、前記パラメータデータの適合指標及びその分布の推定に用いる学習モデルを用いる
　ことを特徴とするフィッティング支援方法。 (Appendix 9)
The fitting support method described in Appendix 8
In the step (b), the machine uses the inspection information of a plurality of users acquired in the past, the survey information of a plurality of users acquired in the past, and the parameter data set in the device in the past as inputs. A fitting support method characterized by using a learning model used for estimating a matching index of the parameter data and its distribution generated by learning.
（付記１０）
　付記９に記載のフィッティング支援方法であって、
（ｄ）フィッティングに用いる前記検査情報、又は前記調査情報、又はそれら両方に不足している情報がある場合、前記不足している情報を補うために、前記不足している情報に関連付けられた質問情報を選択する、ステップを有し、
　前記（ｃ）のステップにおいて、前記質問情報を用いて質問内容を表す前記出力情報を生成して出力装置に出力する
　ことを特徴とするフィッティング支援方法。 (Appendix 10)
The fitting support method described in Appendix 9
(D) If the inspection information used for fitting, the survey information, or both of them have missing information, a question associated with the missing information to supplement the missing information. Select information, have steps,
A fitting support method according to the step (c), wherein the output information representing the question content is generated using the question information and output to an output device.
（付記１１）
　付記１０に記載のフィッティング支援方法であって、
　前記（ｄ）の処理において、前記不足している情報と前記パラメータデータとの相関を表す情報量を、前記不足している情報に対して算出し、算出した前記情報量に基づいて、前記不足している情報を選択する
　ことを特徴とするフィッティング支援方法。 (Appendix 11)
The fitting support method described in Appendix 10
In the process of (d), the amount of information representing the correlation between the deficient information and the parameter data is calculated for the deficient information, and the deficiency is based on the calculated amount of information. A fitting support method characterized by selecting the information to be provided.
（付記１２）
　付記１０又は１１に記載のフィッティング支援方法であって、
　前記（ｂ）の処理において、前記質問情報を用いて前記不足している情報が取得された場合、取得された前記不足している情報を追加して、再度、前記パラメータデータの適合指標及びその分布を推定する
　ことを特徴とするフィッティング支援方法。 (Appendix 12)
The fitting support method according to Appendix 10 or 11.
In the process of (b), when the missing information is acquired using the question information, the acquired information is added, and the conformity index of the parameter data and the matching index thereof are again obtained. A fitting support method characterized by estimating the distribution.
（付記１３）
　コンピュータに、
（ａ）対象者に対して実施した検査の結果を表す検査情報と、前記対象者に関係する調査の結果を表す調査情報とを取得する、ステップと、
（ｂ）前記対象者に機器を適合させるために用いるパラメータデータのフィッティングにおいて、取得した前記検査情報と前記調査情報とを入力して、前記パラメータデータの適合度を表す適合指標及びその分布を推定する、ステップと、
　を実行させる命令を含む、プログラムを記録しているコンピュータ読み取り可能な記録媒体。 (Appendix 13)
On the computer
(A) A step of acquiring inspection information representing the results of an inspection conducted on a subject and survey information representing the results of a survey related to the subject.
(B) In fitting the parameter data used to fit the device to the target person, the acquired inspection information and the survey information are input to estimate the goodness of fit index representing the goodness of fit of the parameter data and its distribution. To do, step and
A computer-readable recording medium that records a program, including instructions to execute.
（付記１４）
　付記１３に記載のコンピュータ読み取り可能な記録媒体であって、
　前記プログラムが、前記コンピュータに、
（ｃ）前記パラメータデータの適合指標及びその分布を表す出力情報を生成して出力装置に出力する、ステップを実行させる命令を更に含む、
　プログラムを記録しているコンピュータ読み取り可能な記録媒体。 (Appendix 14)
The computer-readable recording medium according to Appendix 13.
The program is on the computer
(C) Further including an instruction to execute a step, which generates output information representing the conformity index of the parameter data and its distribution and outputs it to the output device.
A computer-readable recording medium on which the program is recorded.
（付記１５）
　付記１４に記載のコンピュータ読み取り可能な記録媒体であって、
　前記（ｂ）のステップにおいて、過去に取得した複数の利用者の検査情報と、過去に取得した複数の利用者の調査情報と、過去に前記機器に設定されたパラメータデータとを入力として、機械学習により生成された、前記パラメータデータの適合指標及びその分布の推定に用いる学習モデルを用いる
　ことを特徴とするコンピュータ読み取り可能な記録媒体。 (Appendix 15)
The computer-readable recording medium according to Appendix 14.
In the step (b), the machine uses the inspection information of a plurality of users acquired in the past, the survey information of a plurality of users acquired in the past, and the parameter data set in the device in the past as inputs. A computer-readable recording medium generated by learning that uses a learning model used for estimating the conformity index of the parameter data and its distribution.
（付記１６）
　付記１５に記載のコンピュータ読み取り可能な記録媒体であって、
　前記プログラムが、前記コンピュータに、
（ｄ）フィッティングに用いる前記検査情報、又は前記調査情報、又はそれら両方に不足している情報がある場合、前記不足している情報を補うために、前記不足している情報に関連付けられた質問情報を選択する、ステップを実行させる命令を更に含み、
　前記（ｃ）のステップにおいて、前記質問情報を用いて質問内容を表す前記出力情報を生成して出力装置に出力する
　ことを特徴とするコンピュータ読み取り可能な記録媒体。 (Appendix 16)
The computer-readable recording medium according to Appendix 15.
The program is on the computer
(D) If the inspection information used for fitting, the survey information, or both of them have missing information, a question associated with the missing information to supplement the missing information. Includes additional instructions to select information, perform steps,
A computer-readable recording medium, characterized in that, in the step (c), the question information is used to generate the output information representing the question content and output it to an output device.
（付記１７）
　付記１６に記載のコンピュータ読み取り可能な記録媒体であって、
　前記（ｄ）のステップにおいて、前記不足している情報と前記パラメータデータとの相関を表す情報量を、前記不足している情報に対して算出し、算出した前記情報量に基づいて、前記不足している情報を選択する
　ことを特徴とするコンピュータ読み取り可能な記録媒体。 (Appendix 17)
The computer-readable recording medium according to Appendix 16.
In the step (d), an amount of information representing the correlation between the missing information and the parameter data is calculated for the missing information, and based on the calculated information amount, the missing information is obtained. A computer-readable recording medium characterized by selecting information that is being used.
（付記１８）
　付記１６又は１７に記載のコンピュータ読み取り可能な記録媒体であって、
　前記（ｂ）のステップにおいて、前記質問情報を用いて前記不足している情報が取得された場合、取得された前記不足している情報を追加して、再度、前記パラメータデータの適合指標及びその分布を推定する
　ことを特徴とするコンピュータ読み取り可能な記録媒体。 (Appendix 18)
A computer-readable recording medium according to Appendix 16 or 17.
When the missing information is acquired using the question information in the step (b), the acquired information is added to the matching index of the parameter data and the matching index thereof. A computer-readable recording medium characterized by estimating the distribution.
　以上、実施の形態を参照して本願発明を説明したが、本願発明は上記実施の形態に限定されるものではない。本願発明の構成や詳細には、本願発明のスコープ内で当業者が理解し得る様々な変更をすることができる。 Although the present invention has been described above with reference to the embodiments, the present invention is not limited to the above embodiments. Various changes that can be understood by those skilled in the art can be made within the scope of the present invention in terms of the structure and details of the present invention.
　以上のように本発明によれば、フィッティングの精度を向上させることができる。本発明は、生体に装着して用いる機器などを対象者に適合させるフィッティングが必要な分野において有用である。具体的には、機器は、補聴器、イヤホン、スマートグラス、ヘッドマウントディスプレイ、スマートウォッチ、楽器、医療用装具、介護ロボットなどが考えられる。また、機器は、生体に装着せずに用いる機器でもよい。 As described above, according to the present invention, the accuracy of fitting can be improved. The present invention is useful in a field where fitting such as a device worn on a living body to be adapted to a subject is required. Specifically, the device may be a hearing aid, an earphone, a smart glass, a head-mounted display, a smart watch, a musical instrument, a medical device, a nursing robot, or the like. Further, the device may be a device used without being attached to a living body.
　　１　フィッティング支援装置
　　２　取得部
　　３　推定部
　２１　入力装置
　２２　出力装置
　２４　選択部
　２５　出力情報生成部
　２６、２７、２８　学習モデル
　３１　学習装置
　３２　記憶装置
　３３　取得部
　３４、３５　分類部
　３６　生成部
１１０　コンピュータ
１１１　ＣＰＵ
１１２　メインメモリ
１１３　記憶装置
１１４　入力インターフェイス
１１５　表示コントローラ
１１６　データリーダ／ライタ
１１７　通信インターフェイス
１１８　入力機器
１１９　ディスプレイ装置
１２０　記録媒体
１２１　バス 1 Fitting support device 2 Acquisition unit 3 Estimating unit 21 Input device 22 Output device 24 Selection unit 25 Output information generation unit 26, 27, 28 Learning model 31 Learning device 32 Storage device 33 Acquisition unit 34, 35 Classification unit 36 Generation unit 110 Computer 111 CPU
112 Main memory 113 Storage device 114 Input interface 115 Display controller 116 Data reader / writer 117 Communication interface 118 Input device 119 Display device 120 Recording medium 121 Bus

Claims (18)

  1. 　対象者に対して実施した検査の結果を表す検査情報と、前記対象者に関係する調査の結果を表す調査情報とを取得する、取得手段と、
    　前記対象者に機器を適合させるために用いるパラメータデータのフィッティングにおいて、取得した前記検査情報と前記調査情報とを入力して、前記パラメータデータの適合度を表す適合指標及びその分布を推定する、推定手段と、
    　を有することを特徴とするフィッティング支援装置。     An acquisition means for acquiring inspection information representing the results of an inspection conducted on a subject and survey information representing the results of a survey related to the subject.
    In the fitting of the parameter data used to fit the device to the target person, the acquired inspection information and the survey information are input to estimate the goodness of fit index representing the goodness of fit of the parameter data and its distribution. Means and
    A fitting support device characterized by having.
  2. 　請求項１に記載のフィッティング支援装置であって、
    　前記パラメータデータの適合指標及びその分布を表す出力情報を生成して出力装置に出力する、出力情報生成手段と、
    　を有することを特徴とするフィッティング支援装置。     The fitting support device according to claim 1.
    An output information generating means that generates output information representing the matching index of the parameter data and its distribution and outputs the output information to the output device.
    A fitting support device characterized by having.
  3. 　請求項２に記載のフィッティング支援装置であって、
    　前記推定手段は、過去に取得した複数の利用者の検査情報と、過去に取得した複数の利用者の調査情報と、過去に前記機器に設定されたパラメータデータとを入力として、機械学習により生成された、前記パラメータデータの適合指標及びその分布の推定に用いる学習モデルを有する
    　ことを特徴とするフィッティング支援装置。     The fitting support device according to claim 2.
    The estimation means is generated by machine learning by inputting inspection information of a plurality of users acquired in the past, survey information of a plurality of users acquired in the past, and parameter data set in the device in the past. A fitting support device having a learning model used for estimating a matching index of the parameter data and its distribution.
  4. 　請求項３に記載のフィッティング支援装置であって、
    　フィッティングに用いる前記検査情報、又は前記調査情報、又はそれら両方に不足している情報がある場合、前記不足している情報を補うために、前記不足している情報に関連付けられた質問情報を選択する、選択手段と、
    　前記出力情報生成手段は、前記質問情報を用いて質問内容を表す前記出力情報を生成して出力装置に出力する
    　ことを特徴とするフィッティング支援装置。     The fitting support device according to claim 3.
    If the inspection information used for fitting, the survey information, or both of them have missing information, the question information associated with the missing information is selected to make up for the missing information. The means of choice and
    The output information generation means is a fitting support device characterized in that the output information representing the question content is generated using the question information and output to the output device.
  5. 　請求項４に記載のフィッティング支援装置であって、
    　前記選択手段は、前記不足している情報と前記パラメータデータとの相関を表す情報量を、前記不足している情報に対して算出し、算出した前記情報量に基づいて、前記不足している情報を選択する
    　ことを特徴とするフィッティング支援装置。     The fitting support device according to claim 4.
    The selection means calculates an amount of information representing the correlation between the deficient information and the parameter data with respect to the deficient information, and based on the calculated amount of information, the deficient information. A fitting support device characterized by selecting information.
  6. 　請求項４又は５に記載のフィッティング支援装置であって、
    　前記推定手段は、前記質問情報を用いて前記不足している情報が取得された場合、取得された前記不足している情報を追加して、再度、前記パラメータデータの適合指標及びその分布を推定する
    　を有することを特徴とするフィッティング支援装置。     The fitting support device according to claim 4 or 5.
    When the deficient information is acquired by using the question information, the estimation means adds the acquired deficient information and estimates the matching index of the parameter data and its distribution again. A fitting support device characterized by having a device.
  7. （ａ）対象者に対して実施した検査の結果を表す検査情報と、前記対象者に関係する調査の結果を表す調査情報とを取得し、
    （ｂ）前記対象者に機器を適合させるために用いるパラメータデータのフィッティングにおいて、取得した前記検査情報と前記調査情報とを入力して、前記パラメータデータの適合度を表す適合指標及びその分布を推定する
    　ことを特徴とするフィッティング支援方法。     (A) Obtain the inspection information indicating the result of the inspection conducted on the subject and the survey information representing the result of the investigation related to the subject.
    (B) In fitting the parameter data used to fit the device to the target person, the acquired inspection information and the survey information are input to estimate the goodness of fit index representing the goodness of fit of the parameter data and its distribution. A fitting support method characterized by doing.
  8. 　請求項７に記載のフィッティング支援方法であって、
    （ｃ）前記パラメータデータの適合指標及びその分布を表す出力情報を生成して出力装置に出力する、
    　ことを特徴とするフィッティング支援方法。     The fitting support method according to claim 7.
    (C) Generate output information representing the conformity index of the parameter data and its distribution and output it to the output device.
    A fitting support method characterized by that.
  9. 　請求項８に記載のフィッティング支援方法であって、
    　前記（ｂ）の処理において、過去に取得した複数の利用者の検査情報と、過去に取得した複数の利用者の調査情報と、過去に前記機器に設定されたパラメータデータとを入力として、機械学習により生成された、前記パラメータデータの適合指標及びその分布の推定に用いる学習モデルを有する
    　ことを特徴とするフィッティング支援方法。     The fitting support method according to claim 8.
    In the process (b), the machine receives inspection information of a plurality of users acquired in the past, survey information of a plurality of users acquired in the past, and parameter data set in the device in the past as inputs. A fitting support method characterized by having a learning model used for estimating a matching index of the parameter data and its distribution generated by learning.
  10. 　請求項９に記載のフィッティング支援方法であって、
    （ｄ）フィッティングに用いる前記検査情報、又は前記調査情報、又はそれら両方に不足している情報がある場合、前記不足している情報を補うために、前記不足している情報に関連付けられた質問情報を選択し、
    　前記（ｃ）の処理において、前記質問情報を用いて質問内容を表す前記出力情報を生成して出力装置に出力する
    　ことを特徴とするフィッティング支援方法。     The fitting support method according to claim 9.
    (D) If the inspection information used for fitting, the survey information, or both of them have missing information, a question associated with the missing information to supplement the missing information. Select information,
    A fitting support method according to the process (c), wherein the output information representing the question content is generated using the question information and output to an output device.
  11. 　請求項１０に記載のフィッティング支援方法であって、
    　前記（ｄ）の処理において、前記不足している情報と前記パラメータデータとの相関を表す情報量を、前記不足している情報に対して算出し、算出した前記情報量に基づいて、前記不足している情報を選択する
    　ことを特徴とするフィッティング支援方法。     The fitting support method according to claim 10.
    In the process of (d), the amount of information representing the correlation between the deficient information and the parameter data is calculated for the deficient information, and the deficiency is based on the calculated amount of information. A fitting support method characterized by selecting the information to be provided.
  12. 　請求項１０又は１１に記載のフィッティング支援方法であって、
    　前記（ｂ）の処理において、前記質問情報を用いて前記不足している情報が取得された場合、取得された前記不足している情報を追加して、再度、前記パラメータデータの適合指標及びその分布を推定する
    　ことを特徴とするフィッティング支援方法。     The fitting support method according to claim 10 or 11.
    In the process of (b), when the missing information is acquired using the question information, the acquired information is added, and the conformity index of the parameter data and the matching index thereof are again obtained. A fitting support method characterized by estimating the distribution.
  13. 　コンピュータに、
    （ａ）対象者に対して実施した検査の結果を表す検査情報と、前記対象者に関係する調査の結果を表す調査情報とを取得する、ステップと、
    （ｂ）前記対象者に機器を適合させるために用いるパラメータデータのフィッティングにおいて、取得した前記検査情報と前記調査情報とを入力して、前記パラメータデータの適合度を表す適合指標及びその分布を推定する、ステップと、
    　を実行させる命令を含む、プログラムを記録しているコンピュータ読み取り可能な記録媒体。     On the computer
    (A) A step of acquiring inspection information representing the results of an inspection conducted on a subject and survey information representing the results of a survey related to the subject.
    (B) In fitting the parameter data used to fit the device to the target person, the acquired inspection information and the survey information are input to estimate the goodness of fit index representing the goodness of fit of the parameter data and its distribution. To do, step and
    A computer-readable recording medium that records a program, including instructions to execute.
  14. 　請求項１３に記載のコンピュータ読み取り可能な記録媒体であって、
    　前記プログラムが、前記コンピュータに、
    （ｃ）前記パラメータデータの適合指標及びその分布を表す出力情報を生成して出力装置に出力する、ステップを実行させる命令を更に含む、
    　プログラムを記録しているコンピュータ読み取り可能な記録媒体。     The computer-readable recording medium according to claim 13.
    The program is on the computer
    (C) Further including an instruction to execute a step, which generates output information representing the conformity index of the parameter data and its distribution and outputs it to the output device.
    A computer-readable recording medium on which the program is recorded.
  15. 　請求項１４に記載のコンピュータ読み取り可能な記録媒体であって、
    　前記（ｂ）のステップにおいて、過去に取得した複数の利用者の検査情報と、過去に取得した複数の利用者の調査情報と、過去に前記機器に設定されたパラメータデータとを入力として、機械学習により生成された、前記パラメータデータの適合指標及びその分布の推定に用いる学習モデルを有する
    　ことを特徴とするコンピュータ読み取り可能な記録媒体。     The computer-readable recording medium according to claim 14.
    In the step (b), the machine uses the inspection information of a plurality of users acquired in the past, the survey information of a plurality of users acquired in the past, and the parameter data set in the device in the past as inputs. A computer-readable recording medium having a learning model used for estimating a conformity index of the parameter data and its distribution generated by learning.
  16. 　請求項１５に記載のコンピュータ読み取り可能な記録媒体であって、
    　前記プログラムが、前記コンピュータに、
    （ｄ）フィッティングに用いる前記検査情報、又は前記調査情報、又はそれら両方に不足している情報がある場合、前記不足している情報を補うために、前記不足している情報に関連付けられた質問情報を選択する、ステップを実行させる命令を更に含み、
    　前記（ｃ）のステップにおいて、前記質問情報を用いて質問内容を表す前記出力情報を生成して出力装置に出力する
    　ことを特徴とするコンピュータ読み取り可能な記録媒体。     The computer-readable recording medium according to claim 15.
    The program is on the computer
    (D) If the inspection information used for fitting, the survey information, or both of them have missing information, a question associated with the missing information to supplement the missing information. Includes additional instructions to select information, perform steps,
    A computer-readable recording medium, characterized in that, in the step (c), the question information is used to generate the output information representing the question content and output it to an output device.
  17. 　請求項１６に記載のコンピュータ読み取り可能な記録媒体であって、
    　前記（ｄ）のステップにおいて、前記不足している情報と前記パラメータデータとの相関を表す情報量を、前記不足している情報に対して算出し、算出した前記情報量に基づいて、前記不足している情報を選択する
    　ことを特徴とするコンピュータ読み取り可能な記録媒体。     The computer-readable recording medium according to claim 16.
    In the step (d), an amount of information representing the correlation between the missing information and the parameter data is calculated for the missing information, and based on the calculated information amount, the missing information is obtained. A computer-readable recording medium characterized by selecting information that is being used.
  18. 　請求項１６又は１７に記載のコンピュータ読み取り可能な記録媒体であって、
    　前記（ｂ）のステップにおいて、前記質問情報を用いて前記不足している情報が取得された場合、取得された前記不足している情報を追加して、再度、前記パラメータデータの適合指標及びその分布を推定する
    　ことを特徴とするコンピュータ読み取り可能な記録媒体。     A computer-readable recording medium according to claim 16 or 17.
    When the missing information is acquired using the question information in the step (b), the acquired information is added to the matching index of the parameter data and the matching index thereof. A computer-readable recording medium characterized by estimating the distribution.
PCT/JP2019/018077 2019-04-26 2019-04-26 Fitting assistance device, fitting assistance method, and computer-readable recording medium WO2020217494A1 (en)

Priority Applications (2)

Application Number Priority Date Filing Date Title
PCT/JP2019/018077 WO2020217494A1 (en) 2019-04-26 2019-04-26 Fitting assistance device, fitting assistance method, and computer-readable recording medium
JP2021515724A JP7276433B2 (en) 2019-04-26 2019-04-26 FITTING ASSIST DEVICE, FITTING ASSIST METHOD, AND PROGRAM

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
PCT/JP2019/018077 WO2020217494A1 (en) 2019-04-26 2019-04-26 Fitting assistance device, fitting assistance method, and computer-readable recording medium

Publications (1)

Publication Number Publication Date
WO2020217494A1 true WO2020217494A1 (en) 2020-10-29

Family

ID=72940910

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/JP2019/018077 WO2020217494A1 (en) 2019-04-26 2019-04-26 Fitting assistance device, fitting assistance method, and computer-readable recording medium

Country Status (2)

Country Link
JP (1) JP7276433B2 (en)
WO (1) WO2020217494A1 (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20220217486A1 (en) * 2021-01-04 2022-07-07 Gn Hearing A/S Usability and satisfaction of a hearing aid

Citations (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2017152865A (en) * 2016-02-23 2017-08-31 リオン株式会社 Hearing aid fitting device, hearing aid fitting program, hearing aid fitting server, and hearing aid fitting method

Family Cites Families (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP4516366B2 (en) 2004-07-09 2010-08-04 株式会社ビジョンメガネ Hearing test device, hearing aid selection device using the same, management server, and computer-readable recording medium

Patent Citations (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2017152865A (en) * 2016-02-23 2017-08-31 リオン株式会社 Hearing aid fitting device, hearing aid fitting program, hearing aid fitting server, and hearing aid fitting method

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20220217486A1 (en) * 2021-01-04 2022-07-07 Gn Hearing A/S Usability and satisfaction of a hearing aid
US11849288B2 (en) * 2021-01-04 2023-12-19 Gn Hearing A/S Usability and satisfaction of a hearing aid

Also Published As

Publication number Publication date
JP7276433B2 (en) 2023-05-18
JPWO2020217494A1 (en) 2020-10-29

Similar Documents

Publication Publication Date Title
CN109600699B (en) System for processing service request, method and storage medium thereof
US11602287B2 (en) Automatically aiding individuals with developing auditory attention abilities
JPWO2019176986A1 (en) Signal processing systems, signal processing equipment, signal processing methods, and programs
EP4179533A1 (en) Pressure maintenance in an electrolysis system
WO2020212404A1 (en) Hearing test system
US20100027800A1 (en) Automatic Performance Optimization for Perceptual Devices
WO2020217494A1 (en) Fitting assistance device, fitting assistance method, and computer-readable recording medium
US11830478B2 (en) Learning device, learning method, and learning program for images and sound which uses a similarity matrix
JP7272425B2 (en) FITTING ASSIST DEVICE, FITTING ASSIST METHOD, AND PROGRAM
TW202223804A (en) Electronic resource pushing method and system
KR20230043080A (en) Method for screening psychiatric disorder based on voice and apparatus therefor
US20230081796A1 (en) Managing audio content delivery
JP2023027697A (en) Terminal device, transmission method, transmission program and information processing system
JP7307507B2 (en) Pathological condition analysis system, pathological condition analyzer, pathological condition analysis method, and pathological condition analysis program
JP2020017104A (en) Learning device, learning method, and computer program
KR102093365B1 (en) Control method, device, and program of the trial management data-based hearing aid compliance management system
JP2024504435A (en) Audio signal generation system and method
US20240007807A1 (en) System and method for performing consumer hearing aid fittings
US12033614B2 (en) System and method for secure data augmentation for speech processing systems
Cauchi et al. Hardware/software architecture for services in the hearing aid industry
US20240221722A1 (en) Eyewear device and method of use
JP6890867B1 (en) Evaluation program and evaluation system
WO2022264535A1 (en) Information processing method and information processing system
US20240121560A1 (en) Facilitating hearing device fitting
US20230104178A1 (en) System and method for performing consumer hearing aid fittings

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 19925738

Country of ref document: EP

Kind code of ref document: A1

ENP Entry into the national phase

Ref document number: 2021515724

Country of ref document: JP

Kind code of ref document: A

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 19925738

Country of ref document: EP

Kind code of ref document: A1