CN110610320A - Financial risk level prediction method, device, electronic equipment and storage medium - Google Patents
Financial risk level prediction method, device, electronic equipment and storage medium Download PDFInfo
- Publication number
- CN110610320A CN110610320A CN201910885359.4A CN201910885359A CN110610320A CN 110610320 A CN110610320 A CN 110610320A CN 201910885359 A CN201910885359 A CN 201910885359A CN 110610320 A CN110610320 A CN 110610320A
- Authority
- CN
- China
- Prior art keywords
- driver
- risk level
- financial risk
- prediction model
- prediction
- Prior art date
- Legal status (The legal status 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 status listed.)
- Withdrawn
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0635—Risk analysis of enterprise or organisation activities
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q40/00—Finance; Insurance; Tax strategies; Processing of corporate or income taxes
- G06Q40/02—Banking, e.g. interest calculation or account maintenance
Landscapes
- Business, Economics & Management (AREA)
- Engineering & Computer Science (AREA)
- Human Resources & Organizations (AREA)
- Economics (AREA)
- Strategic Management (AREA)
- General Business, Economics & Management (AREA)
- Theoretical Computer Science (AREA)
- Development Economics (AREA)
- Entrepreneurship & Innovation (AREA)
- Physics & Mathematics (AREA)
- Finance (AREA)
- General Physics & Mathematics (AREA)
- Marketing (AREA)
- Accounting & Taxation (AREA)
- Educational Administration (AREA)
- Technology Law (AREA)
- Game Theory and Decision Science (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Tourism & Hospitality (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention provides a financial risk grade prediction method, a device, electronic equipment and a storage medium, wherein the financial risk grade prediction model establishing method comprises the following steps: acquiring driver information and a driver financial risk level; obtaining driver characteristics; forming a driver sample, and forming a driver sample set according to a plurality of driver samples; dividing a driver sample set into N training sets; for each value of i from 1 to N, for the ith training set: modeling the training set by using a machine learning algorithm by taking the average absolute error as a first loss function to obtain an i1 th prediction model; modeling the training set by using a machine learning algorithm by taking the mean square error as a second loss function to obtain an i2 th prediction model; the ith 1 prediction model and the ith 2 prediction model form an ith group of prediction models; and taking the obtained N groups of prediction models as financial risk level prediction models. The method and the device provided by the invention realize effective financial risk level prediction.
Description
Technical Field
The invention relates to the technical field of big data, in particular to a financial risk grade prediction method and device, electronic equipment and a storage medium.
Background
The general finance requires the mining of credit for more people. The current credit investigation system which really plays a role in China is mainly a credit investigation system at the center row, and the covered population is very limited and is far lower than the coverage of the American credit investigation system to 85 percent of the population. The vast number of people in China who are not covered by the traditional credit investigation system also need credit service and enjoy financial prosperity, so that a new idea of credit investigation is needed. With the continuous development of the internet and economy, the rise of internet finance also promotes the rapid development of the credit industry. Therefore, it is clearly of great importance to provide a reliable risk identification scheme. The personal credit assessment technology for serving the internet financial business is generated by depending on mass data.
The patent with publication number CN 110033159 a obtains the relevant data of the enterprise from the official organization, and then extracts the operation behavior feature, revenue situation feature, loan behavior feature, asset variation feature, and administrative reward and punishment feature by using the data, inputs the data to the risk identification model, and outputs the identification result indicating the risk situation of the enterprise by the risk identification model. The method aims at enterprise users, but not individual users, and the individual users have great difference from the enterprise risk level judgment method due to individual difference and behavior complexity, so the enterprise risk level judgment method is not completely suitable for individuals.
The patent with publication number CN 110009479 a obtains credit data of a user and an application service of the user, and checks anti-fraud behavior of the user according to the credit data and anti-fraud rules. And if the user checks the cheating behavior, performing group cheating behavior checking on the user according to the credit data and the group cheating model. The method mainly judges the risk level of the user by judging whether the user is suspected to be fraudulent or not, and has low reliability and interpretability.
However, since the financial risk levels based on the big data are different in data object, data source and judgment criterion, it is difficult to provide a highly reliable and interpretable method for predicting financial risk levels suitable for individuals.
Currently, with the development of enterprises of internet and information service in the logistics industry, logistics transportation platforms based on vehicle and goods matching are produced. On the premise that the logistics transportation platform generates driver related information, how to provide safe and effective financial loan services for truck drivers becomes a key problem to be solved in logistics finance.
Disclosure of Invention
The present invention is directed to a method, an apparatus, an electronic device, and a storage medium for predicting a financial risk level that overcome the limitations and disadvantages of the related art, and thereby overcome one or more of the problems due to the limitations and disadvantages of the related art.
According to one aspect of the invention, a financial risk level prediction model establishing method is provided, which comprises the following steps:
acquiring driver information and a driver financial risk level;
performing feature extraction on the acquired driver information to obtain driver features;
forming a driver sample according to the driver characteristics of a driver and the financial risk level of the driver, and forming a driver sample set according to a plurality of driver samples;
dividing the driver sample set into N training sets, wherein N is an integer greater than or equal to 2;
for each value of i from 1 to N, for the ith training set:
modeling the training set by using a machine learning algorithm by taking the average absolute error as a first loss function to obtain an i1 th prediction model;
modeling the training set by using a machine learning algorithm by taking the mean square error as a second loss function to obtain an i2 th prediction model;
the ith 1 prediction model and the ith 2 prediction model form an ith group of prediction models;
and taking the obtained N groups of prediction models as the financial risk level prediction model.
In some embodiments of the present invention, the extracting the feature of the acquired driver information to obtain the driver feature includes:
for each driver characteristic, a correlation coefficient ρ (a, B) between the driver characteristic and other driver characteristics is calculated as follows:
wherein cov (A, B) represents the covariance of the driver characteristic with other driver characteristics, σ A and σ B represent the standard deviation of the driver characteristic with other driver characteristics;
if the correlation coefficient between the driver characteristic and each of the other driver characteristics is less than the correlation threshold, the driver characteristic is deleted.
In some embodiments of the invention, the first loss function is:
where y is the actual value of the driver's financial risk level,and (3) the model prediction value of the driver financial risk level is obtained, n is the number of samples in the training set, and n is an integer greater than or equal to 2.
In some embodiments of the invention, the second loss function is:
where y is the actual value of the driver's financial risk level,and (3) the model prediction value of the driver financial risk level is obtained, n is the number of samples in the training set, and n is an integer greater than or equal to 2.
In some embodiments of the invention, the machine learning algorithm is a lightweight gradient elevator algorithm.
In some embodiments of the invention, the driver information comprises one or more of driver identity information, vehicle information, device terminal information, transaction payment information, order information, vehicle live information.
According to another aspect of the present invention, there is also provided a financial risk level prediction method, including:
acquiring driver information of a driver;
performing feature extraction on the acquired driver information to obtain driver features;
taking the characteristics of the driver as the input of the financial risk level prediction model, wherein the financial risk level prediction model is established according to the financial risk level prediction model establishing method;
for each value of i from 1 to N, the prediction model of the ith group is:
inputting the driver characteristics into an i1 th prediction model in the N groups of prediction models to obtain an i1 th prediction value;
inputting the driver characteristics into an i2 th prediction model in the N groups of prediction models to obtain an i2 th prediction value;
obtaining a first predicted value according to the N ith 1 predicted values, and obtaining a second predicted value according to the N ith 2 predicted values;
sequencing a plurality of drivers according to the first predicted value;
for a driver positioned at the front a% and the rear b% in the sorting order, the predicted value of the financial risk level of the driver is the weighted sum of the first predicted value and the second predicted value of the driver;
for drivers located at the rest of the ranking order, the predicted value of the driver's financial risk level is the first predicted value of the driver.
According to another aspect of the present invention, there is also provided a financial risk level prediction model creation apparatus, including:
the first acquisition module is used for acquiring driver information and the financial risk level of the driver;
the first feature extraction module is used for carrying out feature extraction on the acquired driver information to obtain driver features;
the sample forming module is used for forming a driver sample according to the driver characteristics of a driver and the financial risk level of the driver, and forming a driver sample set according to a plurality of driver samples;
the dividing module is used for dividing the driver sample set into N training sets, wherein N is an integer greater than or equal to 2;
and the model establishing module is used for performing the following operation on the ith training set for each value from 1 to N of i:
modeling the training set by using a machine learning algorithm by taking the average absolute error as a first loss function to obtain an i1 th prediction model;
modeling the training set by using a machine learning algorithm by taking the mean square error as a second loss function to obtain an i2 th prediction model;
the ith 1 prediction model and the ith 2 prediction model form an ith group of prediction models;
and the prediction model establishing module is used for taking the obtained N groups of prediction models as the financial risk level prediction model.
According to still another aspect of the present invention, there is also provided a financial risk level prediction apparatus including:
the second acquisition module is used for acquiring driver information of a driver;
the second feature extraction module is used for carrying out feature extraction on the acquired driver information to obtain driver features;
an input formation module for inputting the driver characteristics as the input of the financial risk level prediction model, which is established according to the financial risk level prediction model establishment method according to any one of claims 1 to 6;
the predicted value obtaining module is used for obtaining the value of each value from 1 to N of the i, and for the i-th group of prediction models:
inputting the driver characteristics into an i1 th prediction model in the N groups of prediction models to obtain an i1 th prediction value;
inputting the driver characteristics into an i2 th prediction model in the N groups of prediction models to obtain an i2 th prediction value;
obtaining a first predicted value according to the N ith 1 predicted values, and obtaining a second predicted value according to the N ith 2 predicted values;
the sequencing module is used for sequencing a plurality of drivers according to the first prediction value;
a predicted value correction module to:
for a driver positioned at the front a% and the rear b% in the sorting order, the predicted value of the financial risk level of the driver is the weighted sum of the first predicted value and the second predicted value of the driver;
for drivers located at the rest of the ranking order, the predicted value of the driver's financial risk level is the first predicted value of the driver.
According to still another aspect of the present invention, there is also provided an electronic apparatus, including: a processor; a storage medium having stored thereon a computer program which, when executed by the processor, performs the steps as described above.
According to yet another aspect of the present invention, there is also provided a storage medium having stored thereon a computer program which, when executed by a processor, performs the steps as described above.
Compared with the prior art, the invention has the advantages that:
the method can more stably improve the regression precision by performing feature extraction and cross validation on the driver information generated by the logistics transportation platform and applying the combination of a machine learning algorithm, Mean Square Error (MSE) and Mean Absolute Error (MAE) loss function, and can predict the financial risk grade score of the truck driver. The financial risk grade score appears, has strengthened the platform to the management and control of truck driver to and reduce the risk of lending user reimbursement. Meanwhile, the credit system helps vast high-quality driver users with incomplete credit information to obtain financial services more conveniently and quickly.
Drawings
The above and other features and advantages of the present invention will become more apparent by describing in detail exemplary embodiments thereof with reference to the attached drawings.
FIG. 1 shows a flow diagram of a method for establishing a financial risk level prediction model according to an embodiment of the invention.
FIG. 2 shows a flow diagram of a method of financial risk level prediction according to an embodiment of the invention.
Fig. 3 is a block diagram illustrating a financial risk level prediction model creation apparatus according to an embodiment of the present invention.
Fig. 4 shows a block diagram of a financial risk level prediction apparatus according to an embodiment of the present invention.
Fig. 5 schematically illustrates a computer-readable storage medium in an exemplary embodiment of the invention.
Fig. 6 schematically shows an electronic device in an exemplary embodiment of the invention.
Detailed Description
Example embodiments will now be described more fully with reference to the accompanying drawings. Example embodiments may, however, be embodied in many different forms and should not be construed as limited to the examples set forth herein; rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the concept of example embodiments to those skilled in the art. The described features, structures, or characteristics may be combined in any suitable manner in one or more embodiments.
Furthermore, the drawings are merely schematic illustrations of the invention and are not necessarily drawn to scale. The same reference numerals in the drawings denote the same or similar parts, and thus their repetitive description will be omitted. Some of the block diagrams shown in the figures are functional entities and do not necessarily correspond to physically or logically separate entities. These functional entities may be implemented in the form of software, or in one or more hardware modules or integrated circuits, or in different networks and/or processor devices and/or microcontroller devices.
The flow charts shown in the drawings are merely illustrative and do not necessarily include all of the steps. For example, some steps may be decomposed, and some steps may be combined or partially combined, so that the actual execution sequence may be changed according to the actual situation.
FIG. 1 shows a flow diagram of a method for establishing a financial risk level prediction model according to an embodiment of the invention. The method for establishing the financial risk level prediction model comprises the following steps:
step S110: acquiring driver information and a driver financial risk level;
step S120: performing feature extraction on the acquired driver information to obtain driver features;
step S130: forming a driver sample according to the driver characteristics of a driver and the financial risk level of the driver, and forming a driver sample set according to a plurality of driver samples;
step S140: dividing the driver sample set into N training sets, wherein N is an integer greater than or equal to 2;
step S150: for each value of i from 1 to N, for the ith training set:
step S151: modeling the training set by using a machine learning algorithm by taking the average absolute error as a first loss function to obtain an i1 th prediction model;
step S152: modeling the training set by using a machine learning algorithm by taking the mean square error as a second loss function to obtain an i2 th prediction model;
step S153: the ith 1 prediction model and the ith 2 prediction model form an ith group of prediction models;
step S160: and taking the obtained N groups of prediction models as the financial risk level prediction model.
In the financial risk level prediction model establishing method provided by the invention, the financial risk level score of the truck driver can be predicted by performing feature extraction and cross validation on the driver information generated by the logistics transportation platform and applying a machine learning algorithm and a Mean Square Error (MSE) and Mean Absolute Error (MAE) loss function combination method, so that the regression precision can be stably improved. The financial risk grade score appears, has strengthened the platform to the management and control of truck driver to and reduce the risk of lending user reimbursement. Meanwhile, the credit system helps vast high-quality driver users with incomplete credit information to obtain financial services more conveniently and quickly.
In various embodiments of the present invention, the driver information may be obtained from the logistics transportation platform in step S110, and the driver financial risk level may be the driver financial risk level of the driver with complete credit information, obtained from the credit card system.
In various embodiments of the present invention, the driver information may include one or more of driver identity information, vehicle information, device terminal information, transaction payment information, order information, vehicle live information. The driver identity information may include one or more of gender, age, location, whether authentication is passed, date and day when authentication is passed, whether authentication is passed on a road transportation vehicle qualifications, and the like. The vehicle information may include one or more of vehicle age, license plate home, vehicle brand, type of drive-ready vehicle, license plate color, nature of use, total mass, certified load, operational status, and the like. The device terminal information may include one or more of APP version, APP usage days, opening times, goods searching times, call making times, and the like. The transaction payment information may include whether one or more of a bank card is bound, an account balance, a number of days of payment of a subscription, and a number of singular of payment of a subscription. The route information may include one or more of a distance route, a feedback rate, and a rate. The order information may include one or more of an order number, a complaint evaluation, a number of blacklists, a number of account abnormalities, and a number of transaction abnormalities. The vehicle presence information may include one or more of brake switch status, whether positioning is on, GPS average speed, GPS top speed, etc.
In various embodiments of the present invention, after the step S110, a step of data preprocessing is further included before the step S120. And preprocessing the data, namely performing interpolation supplement on missing values and removing abnormal points.
In various embodiments of the present invention, the step S120 may include a step of converting the text information into the encoded information. Specifically, the text information (such as the location of the user, the license plate location, the vehicle brand, the quasi-driving vehicle type, and the like) may be converted into the feature code information, and the specific conversion mode may adopt an algorithm such as word2vec, and is not described herein again.
In some embodiments of the present invention, the step S120 of extracting the features of the acquired driver information to obtain the driver features may further include a step of feature screening. For example, the features may be filtered in such a manner that, for each driver feature, a correlation coefficient ρ (a, B) between the driver feature and other driver features is calculated as follows:
wherein cov (A, B) represents the covariance of the driver characteristic with other driver characteristics, σ A and σ B represent the standard deviation of the driver characteristic with other driver characteristics;
if the correlation coefficient between the driver characteristic and each of the other driver characteristics is less than the correlation threshold, the driver characteristic is deleted. The correlation threshold may be set to 0.001, for example, but the invention is not limited thereto.
In some embodiments of the invention, the first loss function is:
where y is the actual value of the driver's financial risk level,and (3) the model prediction value of the driver financial risk level is obtained, n is the number of samples in the training set, and n is an integer greater than or equal to 2.
In some embodiments of the invention, the second loss function is:
where y is the actual value of the driver's financial risk level,and (3) the model prediction value of the driver financial risk level is obtained, n is the number of samples in the training set, and n is an integer greater than or equal to 2.
In some embodiments of the invention, the machine learning algorithm is a lightweight gradient hoist (LightGBM) algorithm.
Thus, in one embodiment, step S140 combines the driver sample set XDivided into N training sets (A)1,A2,A3,A4,…,AN)。
For A1,A2,A3,A4,…,ANEach training set in turn performs the following steps. With A1For example, the LightGBM algorithm is adopted for A1Modeling is carried out to obtain a prediction model11. Wherein the loss function is:
wherein, y is an actual value,and (4) predicting the value of the model.
Using LightGBM algorithm for A1Modeling is carried out to obtain a prediction model12. Wherein the loss function is:
wherein, y is an actual value,and (4) predicting the value of the model.
Analogizing in sequence to obtain N groups of prediction models (models)21，model22)，(model31，model32),…,(modelN1，modelN2) As the financial risk level prediction model.
After the financial risk level prediction model is established, the invention further provides a financial risk level prediction method, and referring to fig. 2, fig. 2 shows a flowchart of the financial risk level prediction method according to the embodiment of the invention. Fig. 2 shows the following steps together:
step S210: acquiring driver information of a driver;
step S220: performing feature extraction on the acquired driver information to obtain driver features;
step S230: taking the characteristics of the driver as the input of the financial risk level prediction model, wherein the financial risk level prediction model is established according to the financial risk level prediction model establishing method;
step S240: for each value of i from 1 to N, the prediction model of the ith group is:
step S241: inputting the driver characteristics into an i1 th prediction model in the N groups of prediction models to obtain an i1 th prediction value;
step S242: inputting the driver characteristics into an i2 th prediction model in the N groups of prediction models to obtain an i2 th prediction value;
step S250: obtaining a first predicted value according to the N ith 1 predicted values, and obtaining a second predicted value according to the N ith 2 predicted values;
step S260: sequencing a plurality of drivers according to the first predicted value;
step S270: for a driver positioned at the front a% and the rear b% in the sorting order, the predicted value of the financial risk level of the driver is the weighted sum of the first predicted value and the second predicted value of the driver; for drivers located at the rest of the ranking order, the predicted value of the driver's financial risk level is the first predicted value of the driver.
In some embodiments of the present invention, step S240 above corresponds to truck driver features for which the financial risk level score is unknown, followed by (model)11，model12)，(model21，model22),…,(modelN1，modelN2) Modeling, yielding a result score (pred)11，pred12)，(pred21，pred22),…,(modelN1，modelN2)。
Step S250 may take an average value of the N ith 1 predicted values as a first predicted value:
taking the average value of the N ith 2 predicted values as a second predicted value:
in some embodiments of the present invention, a and b may be constants of 0 to 20, and a and b may be the same or different, and the present invention is not limited thereto.
In some embodiments, the obtained predicted value may be displayed in different color gradient changes, for example, the lower the predicted value, the more red the color, the higher the predicted value, and the more green the color, and the predicted value may be visually displayed through the change process and change degree of the color from red to green, so that the obtained predicted value is visually displayed based on the color sense of human eyes.
In the financial risk level prediction method provided by the invention, the function of square loss (MSE) has larger punishment to a larger value, the index for optimizing MAE pays more attention to the optimization to a smaller value than the MSE, the MSE has more accurate prediction to extreme values at two ends, and the MAE has more accurate prediction to the middle, so that simple weighting correction is carried out near the extreme value predicted by the function. The MSE squares the error e (let e be true-predicted), so if e>1, the MSE further increases the error. If there are outliers in the data, the value of e is large, and e is large2It will be much larger than | e |. Therefore, a model using MSE gives more weight to outliers than a model using MAE to compute losses. The model for computing the loss using MSE is updated toward reducing outlier errors at the expense of errors in other samples. However, this reduces the overall performance of the model. MAE loss is better used if the training data is contaminated with outliers. MAE is also more stable than MSE for outliers. In the embodiment, the mode of combining MSE and MAE is adopted, so that the regression precision can be stably improved. The financial risk grade score appears, has strengthened the platform to the management and control of truck driver to and reduce the risk of lending user reimbursement. Meanwhile, the method helps vast high-quality driver users with incomplete credit information to apply for financial loan more conveniently and quickly.
The above is merely one or more specific implementations provided for by the present invention, which is not intended to be limiting.
According to another aspect of the present invention, there is also provided a financial risk level prediction model building apparatus, and fig. 3 shows a block diagram of the financial risk level prediction model building apparatus according to the embodiment of the present invention. The financial risk level prediction apparatus 300 includes a first obtaining module 310, a first feature extraction module 320, a sample formation module 330, a division module 340, a model building module 350, and a prediction model building module 360.
The first acquisition module is used for acquiring driver information and the financial risk level of the driver;
the first feature extraction module is used for carrying out feature extraction on the acquired driver information to obtain driver features;
the sample forming module is used for forming a driver sample according to the driver characteristics of a driver and the financial risk level of the driver, and forming a driver sample set according to a plurality of driver samples;
the dividing module is used for dividing the driver sample set into N training sets, wherein N is an integer greater than or equal to 2;
and the model establishing module is used for performing the following operation on the ith training set for each value from 1 to N of i:
modeling the training set by using a machine learning algorithm by taking the average absolute error as a first loss function to obtain an i1 th prediction model;
modeling the training set by using a machine learning algorithm by taking the mean square error as a second loss function to obtain an i2 th prediction model;
the ith 1 prediction model and the ith 2 prediction model form an ith group of prediction models;
and the prediction model establishing module is used for taking the obtained N groups of prediction models as the financial risk level prediction model.
In the financial risk level prediction model establishment device provided by the invention, on one hand, the financial risk level score of a truck driver can be predicted by a method for stably improving the regression precision by performing feature extraction and cross validation on driver information generated by a logistics transportation platform and combining a machine learning algorithm, Mean Square Error (MSE) and an average absolute error (MAE) loss function. The financial risk grade score appears, has strengthened the platform to the management and control of truck driver to and reduce the risk of lending user reimbursement. Meanwhile, the credit system helps vast high-quality driver users with incomplete credit information to obtain financial services more conveniently and quickly.
Fig. 3 is a schematic diagram of the financial risk level prediction model building apparatus 300 provided by the present invention, and the splitting, merging and adding of modules are within the protection scope of the present invention without departing from the concept of the present invention. The financial risk level prediction model building apparatus 300 provided by the present invention may be implemented by software, hardware, firmware, plug-in, and any combination thereof, which is not limited by the present invention.
According to another aspect of the present invention, there is also provided a financial risk level prediction apparatus, and fig. 4 shows a block diagram of the financial risk level prediction apparatus according to an embodiment of the present invention. The financial risk level prediction apparatus 400 includes a second obtaining module 410, a second feature extraction module 420, an input formation module 430, a predicted value obtaining module 440, a ranking module 450, and a predicted value correction module 460.
The second obtaining module 410 is used for obtaining driver information of a driver;
the second feature extraction module 420 is configured to perform feature extraction on the acquired driver information to obtain a driver feature;
an input forming module 430 for inputting the driver characteristics as the financial risk level prediction model, which is built according to the financial risk level prediction model building method of any one of claims 1 to 6;
the predicted value obtaining module 440 is configured to, for each value of i from 1 to N, perform, for the ith group of prediction models:
inputting the driver characteristics into an i1 th prediction model in the N groups of prediction models to obtain an i1 th prediction value;
inputting the driver characteristics into an i2 th prediction model in the N groups of prediction models to obtain an i2 th prediction value;
obtaining a first predicted value according to the N ith 1 predicted values, and obtaining a second predicted value according to the N ith 2 predicted values;
the ranking module 450 is configured to rank the plurality of drivers according to the first predicted value;
the predictor correction module 460 operates to:
for a driver positioned at the front a% and the rear b% in the sorting order, the predicted value of the financial risk level of the driver is the weighted sum of the first predicted value and the second predicted value of the driver; for drivers located at the rest of the ranking order, the predicted value of the driver's financial risk level is the first predicted value of the driver.
In the financial risk level prediction device provided by the invention, the penalty of a square loss (MSE) function for a large value is larger, the optimization of a small value is emphasized more by an index for optimizing MAE than the MSE, the MSE is more accurate for the prediction of extreme values at two ends, and the MAE is more accurate for the prediction in the middle, so that simple weighting correction is carried out near the extreme value predicted by the function. The MSE squares the error e (let e be true-predicted), so if e>1, the MSE further increases the error. If there are outliers in the data, the value of e is large, and e is large2It will be much larger than | e |. Therefore, a model using MSE gives more weight to outliers than a model using MAE to compute losses. The model for computing the loss using MSE is updated toward reducing outlier errors at the expense of errors in other samples. However, this reduces the overall performance of the model. MAE loss is better used if the training data is contaminated with outliers. MAE is also more stable than MSE for outliers. In the embodiment, the mode of combining MSE and MAE is adopted, so that the regression precision can be stably improved. The financial risk grade score appears, has strengthened the platform to the management and control of truck driver to and reduce the risk of lending user reimbursement. Meanwhile, the method helps vast high-quality driver users with incomplete credit information to apply for financial loan more conveniently and quickly.
Fig. 4 is a schematic diagram of the financial risk level prediction apparatus 400 provided by the present invention, and the splitting, combining and adding of modules are within the protection scope of the present invention without departing from the concept of the present invention. The financial risk level prediction apparatus 400 provided by the present invention may be implemented by software, hardware, firmware, plug-in, and any combination thereof, which is not limited by the present invention.
In an exemplary embodiment of the present invention, a computer-readable storage medium is further provided, on which a computer program is stored, which when executed by a processor, for example, can implement the steps of the financial risk level prediction model building method and the financial risk level prediction method in any one of the above embodiments. In some possible embodiments, aspects of the present invention may also be implemented in the form of a program product including program code for causing a terminal device to perform the steps according to various exemplary embodiments of the present invention described in the methods for establishing a financial risk level prediction model and the methods for predicting a financial risk level section described above in this specification when the program product is run on the terminal device.
Referring to fig. 5, a program product 700 for implementing the above method according to an embodiment of the present invention is described, which may employ a portable compact disc read only memory (CD-ROM) and include program code, and may be run on a terminal device, such as a personal computer. However, the program product of the present invention is not limited in this regard and, in the present document, a readable storage medium may be any tangible medium that can contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device.
The program product may employ any combination of one or more readable media. The readable medium may be a readable signal medium or a readable storage medium. A readable storage medium may be, for example, but not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination of the foregoing. More specific examples (a non-exhaustive list) of the readable storage medium include: an electrical connection having one or more wires, a portable disk, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.
The computer readable storage medium may include a propagated data signal with readable program code embodied therein, for example, in baseband or as part of a carrier wave. Such a propagated data signal may take many forms, including, but not limited to, electro-magnetic, optical, or any suitable combination thereof. A readable storage medium may also be any readable medium that is not a readable storage medium and that can communicate, propagate, or transport a program for use by or in connection with an instruction execution system, apparatus, or device. Program code embodied on a readable storage medium may be transmitted using any appropriate medium, including but not limited to wireless, wireline, optical fiber cable, RF, etc., or any suitable combination of the foregoing.
Program code for carrying out operations for aspects of the present invention may be written in any combination of one or more programming languages, including an object oriented programming language such as Java, C + + or the like and conventional procedural programming languages, such as the "C" programming language or similar programming languages. The program code may execute entirely on the tenant computing device, partly on the tenant device, as a stand-alone software package, partly on the tenant computing device and partly on a remote computing device, or entirely on the remote computing device or server. In the case of remote computing devices, the remote computing devices may be connected to the tenant computing device through any kind of network, including a Local Area Network (LAN) or a Wide Area Network (WAN), or may be connected to an external computing device (e.g., through the internet using an internet service provider).
In an exemplary embodiment of the invention, there is also provided an electronic device that may include a processor and a memory for storing executable instructions of the processor. Wherein the processor is configured to execute the executable instructions to perform the steps of the financial risk level prediction model building method and the financial risk level prediction method in any of the above embodiments.
As will be appreciated by one skilled in the art, aspects of the present invention may be embodied as a system, method or program product. Thus, various aspects of the invention may be embodied in the form of: an entirely hardware embodiment, an entirely software embodiment (including firmware, microcode, etc.) or an embodiment combining hardware and software aspects that may all generally be referred to herein as a "circuit," module "or" system.
An electronic device 500 according to this embodiment of the invention is described below with reference to fig. 6. The electronic device 500 shown in fig. 6 is only an example and should not bring any limitation to the functions and the scope of use of the embodiments of the present invention.
As shown in fig. 6, the electronic device 500 is embodied in the form of a general purpose computing device. The components of the electronic device 500 may include, but are not limited to: at least one processing unit 510, at least one memory unit 520, a bus 530 that couples various system components including the memory unit 520 and the processing unit 510, a display unit 540, and the like.
Wherein the storage unit stores program code executable by the processing unit 510 to cause the processing unit 510 to perform the steps according to various exemplary embodiments of the present invention described in the methods for establishing a financial risk level prediction model and methods for predicting a financial risk level section of this specification. For example, the processing unit 510 may perform the steps as shown in fig. 1 or fig. 2.
The memory unit 520 may include a readable medium in the form of a volatile memory unit, such as a random access memory unit (RAM)5201 and/or a cache memory unit 5202, and may further include a read only memory unit (ROM) 5203.
The memory unit 520 may also include a program/utility 5204 having a set (at least one) of program modules 5205, such program modules 5205 including, but not limited to: an operating system, one or more application programs, other program modules, and program data, each of which, or some combination thereof, may comprise an implementation of a network environment.
Bus 530 may be one or more of any of several types of bus structures including a memory unit bus or memory unit controller, a peripheral bus, an accelerated graphics port, a processing unit, or a local bus using any of a variety of bus architectures.
The electronic device 500 may also communicate with one or more external devices 600 (e.g., keyboard, pointing device, bluetooth device, etc.), with one or more devices that enable a tenant to interact with the electronic device 500, and/or with any devices (e.g., router, modem, etc.) that enable the electronic device 500 to communicate with one or more other computing devices. Such communication may occur via input/output (I/O) interfaces 550. Also, the electronic device 500 may communicate with one or more networks (e.g., a Local Area Network (LAN), a Wide Area Network (WAN), and/or a public network, such as the internet) via the network adapter 560. The network adapter 560 may communicate with other modules of the electronic device 500 via the bus 530. It should be appreciated that although not shown in the figures, other hardware and/or software modules may be used in conjunction with the electronic device 500, including but not limited to: microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, and data backup storage systems, among others.
Through the above description of the embodiments, those skilled in the art will readily understand that the exemplary embodiments described herein may be implemented by software, or by software in combination with necessary hardware. Therefore, the technical solution according to the embodiment of the present invention may be embodied in the form of a software product, which may be stored in a non-volatile storage medium (which may be a CD-ROM, a usb disk, a removable hard disk, etc.) or on a network, and includes several instructions to enable a computing device (which may be a personal computer, a server, or a network device, etc.) to execute the above-mentioned financial risk level prediction model building method and financial risk level prediction method according to the embodiment of the present invention.
Compared with the prior art, the invention has the advantages that:
the method can more stably improve the regression precision by performing feature extraction and cross validation on the driver information generated by the logistics transportation platform and applying the combination of a machine learning algorithm, Mean Square Error (MSE) and Mean Absolute Error (MAE) loss function, and can predict the financial risk grade score of the truck driver. The financial risk grade score appears, has strengthened the platform to the management and control of truck driver to and reduce the risk of lending user reimbursement. Meanwhile, the credit system helps vast high-quality driver users with incomplete credit information to obtain financial services more conveniently and quickly.
Other embodiments of the invention will be apparent to those skilled in the art from consideration of the specification and practice of the invention disclosed herein. This application is intended to cover any variations, uses, or adaptations of the invention following, in general, the principles of the invention and including such departures from the present disclosure as come within known or customary practice within the art to which the invention pertains. It is intended that the specification and examples be considered as exemplary only, with a true scope and spirit of the invention being indicated by the following claims.
Claims (11)
1. A financial risk level prediction model building method is characterized by comprising the following steps:
acquiring driver information and a driver financial risk level;
performing feature extraction on the acquired driver information to obtain driver features;
forming a driver sample according to the driver characteristics of a driver and the financial risk level of the driver, and forming a driver sample set according to a plurality of driver samples;
dividing the driver sample set into N training sets, wherein N is an integer greater than or equal to 2;
for each value of i from 1 to N, for the ith training set:
modeling the training set by using a machine learning algorithm by taking the average absolute error as a first loss function to obtain an i1 th prediction model;
modeling the training set by using a machine learning algorithm by taking the mean square error as a second loss function to obtain an i2 th prediction model;
the ith 1 prediction model and the ith 2 prediction model form an ith group of prediction models; and taking the obtained N groups of prediction models as the financial risk level prediction model.
2. The method of claim 1, wherein the step of performing feature extraction on the obtained driver information to obtain the driver features comprises:
for each driver characteristic, a correlation coefficient ρ (a, B) between the driver characteristic and other driver characteristics is calculated as follows:
wherein cov (A, B) represents the covariance of the driver characteristic with other driver characteristics, σ A and σ B represent the standard deviation of the driver characteristic with other driver characteristics;
if the correlation coefficient between the driver characteristic and each of the other driver characteristics is less than the correlation threshold, the driver characteristic is deleted.
3. The method of building a financial risk level prediction model according to claim 1, wherein the first loss function is:
where y is the actual value of the driver's financial risk level,and (3) the model prediction value of the driver financial risk level is obtained, n is the number of samples in the training set, and n is an integer greater than or equal to 2.
4. The method of building a financial risk level prediction model according to claim 1, wherein the second loss function is:
where y is the actual value of the driver's financial risk level,and (3) the model prediction value of the driver financial risk level is obtained, n is the number of samples in the training set, and n is an integer greater than or equal to 2.
5. The method of any one of claims 1 to 4, wherein the machine learning algorithm is a lightweight gradient elevator algorithm.
6. The financial risk level prediction model building method of any one of claims 1 to 4, wherein the driver information includes one or more of driver identity information, vehicle information, equipment terminal information, transaction payment information, order information, vehicle live information.
7. A method for predicting a financial risk level, comprising:
acquiring driver information of a driver;
performing feature extraction on the acquired driver information to obtain driver features;
inputting the driver characteristics as the financial risk level prediction model, which is built according to the financial risk level prediction model building method of any one of claims 1 to 6;
for each value of i from 1 to N, the prediction model of the ith group is:
inputting the driver characteristics into an i1 th prediction model in the N groups of prediction models to obtain an i1 th prediction value;
inputting the driver characteristics into an i2 th prediction model in the N groups of prediction models to obtain an i2 th prediction value;
obtaining a first predicted value according to the N ith 1 predicted values, and obtaining a second predicted value according to the N ith 2 predicted values;
sequencing a plurality of drivers according to the first predicted value;
for a driver positioned at the front a% and the rear b% in the sorting order, the predicted value of the financial risk level of the driver is the weighted sum of the first predicted value and the second predicted value of the driver;
for drivers located at the rest of the ranking order, the predicted value of the driver's financial risk level is the first predicted value of the driver.
8. A financial risk level prediction model building device is characterized by comprising:
the first acquisition module is used for acquiring driver information and the financial risk level of the driver;
the first feature extraction module is used for carrying out feature extraction on the acquired driver information to obtain driver features;
the sample forming module is used for forming a driver sample according to the driver characteristics of a driver and the financial risk level of the driver, and forming a driver sample set according to a plurality of driver samples;
the dividing module is used for dividing the driver sample set into N training sets, wherein N is an integer greater than or equal to 2;
and the model establishing module is used for performing the following operation on the ith training set for each value from 1 to N of i:
modeling the training set by using a machine learning algorithm by taking the average absolute error as a first loss function to obtain an i1 th prediction model;
modeling the training set by using a machine learning algorithm by taking the mean square error as a second loss function to obtain an i2 th prediction model;
the ith 1 prediction model and the ith 2 prediction model form an ith group of prediction models; and the prediction model establishing module is used for taking the obtained N groups of prediction models as the financial risk level prediction model.
9. A financial risk level prediction apparatus, comprising:
the second acquisition module is used for acquiring driver information of a driver;
the second feature extraction module is used for carrying out feature extraction on the acquired driver information to obtain driver features;
an input formation module for inputting the driver characteristics as the input of the financial risk level prediction model, which is established according to the financial risk level prediction model establishment method according to any one of claims 1 to 6;
the predicted value obtaining module is used for obtaining the value of each value from 1 to N of the i, and for the i-th group of prediction models:
inputting the driver characteristics into an i1 th prediction model in the N groups of prediction models to obtain an i1 th prediction value;
inputting the driver characteristics into an i2 th prediction model in the N groups of prediction models to obtain an i2 th prediction value;
obtaining a first predicted value according to the N ith 1 predicted values, and obtaining a second predicted value according to the N ith 2 predicted values;
the sequencing module is used for sequencing a plurality of drivers according to the first prediction value;
a predicted value correction module to:
for a driver positioned at the front a% and the rear b% in the sorting order, the predicted value of the financial risk level of the driver is the weighted sum of the first predicted value and the second predicted value of the driver;
for drivers located at the rest of the ranking order, the predicted value of the driver's financial risk level is the first predicted value of the driver.
10. An electronic device, characterized in that the electronic device comprises:
a processor;
a memory having stored thereon a computer program which, when executed by the processor, performs the method of establishing a financial risk level prediction model according to any one of claims 1 to 6 or the method of predicting a financial risk level according to claim 7.
11. A storage medium having stored thereon a computer program which, when executed by a processor, executes the method for establishing a financial risk level prediction model according to any one of claims 1 to 6 or the method for predicting a financial risk level according to claim 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910885359.4A CN110610320A (en) | 2019-09-18 | 2019-09-18 | Financial risk level prediction method, device, electronic equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910885359.4A CN110610320A (en) | 2019-09-18 | 2019-09-18 | Financial risk level prediction method, device, electronic equipment and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110610320A true CN110610320A (en) | 2019-12-24 |
Family
ID=68892941
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910885359.4A Withdrawn CN110610320A (en) | 2019-09-18 | 2019-09-18 | Financial risk level prediction method, device, electronic equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110610320A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111611010A (en) * | 2020-04-24 | 2020-09-01 | 武汉大学 | Interpretable method for code modification real-time defect prediction |
-
2019
- 2019-09-18 CN CN201910885359.4A patent/CN110610320A/en not_active Withdrawn
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111611010A (en) * | 2020-04-24 | 2020-09-01 | 武汉大学 | Interpretable method for code modification real-time defect prediction |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109102393B (en) | Method and device for training and using relational network embedded model | |
| CN110070391B (en) | Data processing method and device, computer readable medium and electronic equipment | |
| CN111127178A (en) | Data processing method and device, storage medium and electronic equipment | |
| Rashidi et al. | Bus dwell time modeling using gene expression programming | |
| CN109784848B (en) | Hotel order processing method and related product | |
| CN111582645B (en) | APP risk assessment method and device based on factoring machine and electronic equipment | |
| CN112348321A (en) | Risk user identification method and device and electronic equipment | |
| CN113570222A (en) | User equipment identification method and device and computer equipment | |
| CN107798125B (en) | Access judgment method, system, equipment and storage medium based on intimacy model | |
| CN111582649B (en) | Risk assessment method and device based on user APP single-heat coding and electronic equipment | |
| CN110610320A (en) | Financial risk level prediction method, device, electronic equipment and storage medium | |
| CN112488865A (en) | Financial risk prediction method and device based on financial time nodes and electronic equipment | |
| CN110689257B (en) | Operator big data-based fast-moving consumer goods business supervision method and device | |
| CN117437020A (en) | Merchant risk judging method and device, electronic equipment and medium | |
| CN109636627B (en) | Insurance product management method, device, medium and electronic equipment based on block chain | |
| CN111382909A (en) | Rejection inference method based on survival analysis model expansion bad sample and related equipment | |
| CN116664306A (en) | Intelligent recommendation method and device for wind control rules, electronic equipment and medium | |
| CN114742645B (en) | User security level identification method and device based on multi-stage time sequence multitask | |
| CN113487208B (en) | Risk assessment method and risk assessment device | |
| CN116091249A (en) | Transaction risk assessment method, device, electronic equipment and medium | |
| CN109816234A (en) | Service access method, service access device, electronic equipment and storage medium | |
| CN113298637A (en) | User diversion method, device and system of service platform | |
| CN114399224A (en) | Data processing method, device, equipment, medium and product | |
| CN113420789A (en) | Method, device, storage medium and computer equipment for predicting risk account | |
| CN113516398A (en) | Risk equipment identification method and device based on hierarchical sampling and electronic equipment |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| WW01 | Invention patent application withdrawn after publication |
Application publication date: 20191224 |
|
| WW01 | Invention patent application withdrawn after publication |