CN114943307A - Model training method and device, storage medium and electronic equipment - Google Patents
Model training method and device, storage medium and electronic equipment Download PDFInfo
- Publication number
- CN114943307A CN114943307A CN202210752076.4A CN202210752076A CN114943307A CN 114943307 A CN114943307 A CN 114943307A CN 202210752076 A CN202210752076 A CN 202210752076A CN 114943307 A CN114943307 A CN 114943307A
- Authority
- CN
- China
- Prior art keywords
- data set
- classification
- training
- model
- wind control
- 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.)
- Pending
Links
- 238000012549 training Methods 0.000 title claims abstract description 142
- 238000000034 method Methods 0.000 title claims abstract description 61
- 238000003860 storage Methods 0.000 title claims abstract description 20
- 238000013145 classification model Methods 0.000 claims abstract description 110
- 238000009826 distribution Methods 0.000 claims abstract description 22
- 230000000694 effects Effects 0.000 claims description 22
- 238000004590 computer program Methods 0.000 claims description 17
- 238000000605 extraction Methods 0.000 claims description 5
- 230000008569 process Effects 0.000 description 14
- 238000010586 diagram Methods 0.000 description 12
- 230000006870 function Effects 0.000 description 10
- 230000006872 improvement Effects 0.000 description 9
- 238000012545 processing Methods 0.000 description 7
- 238000005516 engineering process Methods 0.000 description 4
- 230000015556 catabolic process Effects 0.000 description 3
- 238000006731 degradation reaction Methods 0.000 description 3
- 230000003287 optical effect Effects 0.000 description 3
- 230000004048 modification Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004075 alteration Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 230000001413 cellular effect Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 238000010924 continuous production Methods 0.000 description 1
- 230000007123 defense Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 229920001296 polysiloxane Polymers 0.000 description 1
- 230000000750 progressive effect Effects 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
- 238000012795 verification Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Evolutionary Computation (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Software Systems (AREA)
- Mathematical Physics (AREA)
- Health & Medical Sciences (AREA)
- Biomedical Technology (AREA)
- Computing Systems (AREA)
- Molecular Biology (AREA)
- General Health & Medical Sciences (AREA)
- Evolutionary Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The specification discloses a method, a device, a storage medium and an electronic device for model training, which are used for collecting service data of online service within set time, constructing a service data set, training a preset classification model according to the service data set and a sample data set used in training a wind control model, wherein the classification model is used for judging which data set the input data belongs to, then determining the training quality degree of the trained classification model, determining the integral discrimination of the service data set and the sample data set according to the training quality degree, and training the wind control model according to the discrimination so as to perform wind control on an online user through the trained wind control model. Therefore, the method indirectly determines how different the sample distribution of the business data set and the sample data set exists through the training quality of the training classification model, and improves the accuracy of wind control through the wind control model.
Description
Technical Field
The present disclosure relates to the field of wind control, and in particular, to a method, an apparatus, a storage medium, and an electronic device for model training.
Background
Currently, the internet platform needs to guarantee information security of a user, and therefore the internet platform needs to perform wind control on the user, and can timely stop a service executed for the user when the service of the user has a risk, or remind the user of paying attention to the service which may have the risk.
In practical application, the internet platform can carry out wind control through a wind control model (machine learning model), but in the wind control business, along with the application of the wind control model for a long time, certain differences can appear in the sample distribution between the samples which are generated on line and used for wind control and the samples which train the wind control model at first, so that the wind control effect of the wind control model can be worse and worse, and thus, the wind control can not be accurately carried out, and the safety of the platform is difficult to guarantee.
Therefore, how to improve the accuracy of the wind control and ensure the safety of the platform is a problem to be solved urgently.
Disclosure of Invention
The specification provides a model training method, a model training device, a storage medium and electronic equipment, so that the effect of a wind control model is prevented from declining, the accuracy of wind control is improved, and the safety of a platform is ensured.
The technical scheme adopted by the specification is as follows:
the present specification provides a method of model training comprising:
acquiring service data of online services within a set time, and constructing a service data set;
training a preset classification model according to the business data set and a sample data set used in training a wind control model, wherein the classification model is used for judging which data set of the business data set and the sample data set the input data belongs to;
determining the training quality degree of the trained classification model;
determining the integral discrimination of the business data set and the sample data set according to the training quality degree;
and training the wind control model according to the discrimination so as to perform wind control on the online user through the trained wind control model.
Optionally, determining a training quality degree of the trained classification model includes:
inputting the business data set and the sample data set into the trained classification model to obtain a classification result; determining the classification accuracy and the classification error rate corresponding to the trained classification model according to the classification result;
determining a classification effect distribution curve corresponding to the trained classification model according to the classification accuracy and the classification error rate;
and determining the training quality degree according to the classification effect distribution curve.
Optionally, inputting the service data set and the sample data set into the trained classification model to obtain a classification result, including:
taking the sample data set as a positive sample set, taking the service data set as a negative sample set, and inputting the positive sample set and the negative sample set into the trained classification model to obtain a classification result for the positive sample set and a classification result for the negative sample set;
according to the classification result, determining the classification accuracy and the classification error rate corresponding to the trained classification model, specifically comprising:
and obtaining the classification correct rate according to the proportion of the correctly identified positive samples in all the positive samples, and obtaining the classification error rate according to the proportion of the incorrectly identified negative samples as the positive samples in all the negative samples.
Optionally, training the wind control model according to the discrimination includes:
and if the discrimination is not smaller than the set discrimination, training the wind control model according to the service data set.
Optionally, before training a preset classification model according to the service data set and the training data set, the method further includes:
and constructing the classification model according to the model structure corresponding to the wind control model.
Optionally, constructing the classification model according to a model structure corresponding to the wind control model, including:
constructing the classification model according to a specified network layer in a model structure corresponding to the wind control model, wherein the specified network layer comprises: and the characteristic extraction layer is used for inputting the data characteristic dimension of the data of the wind control model to be the same as the data characteristic dimension of the data of the classification model.
The present specification provides an apparatus for model training, comprising:
the acquisition module is used for acquiring service data of online services in set time and constructing a service data set;
the training module is used for training a preset classification model according to the business data set and a sample data set used when the wind control model is trained, and the classification model is used for judging which data set of the business data set and the sample data set the input data belongs to;
the first determining module is used for determining the training quality degree of the trained classification model;
the second determining module is used for determining the integral discrimination of the business data set and the sample data set according to the training quality degree;
and the wind control module is used for training the wind control model according to the discrimination so as to perform wind control on the online user through the trained wind control model.
The present specification provides a computer-readable storage medium storing a computer program which, when executed by a processor, implements the above-described method of model training.
The present specification provides an electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, the processor implementing the above-described method of model training when executing the program.
The technical scheme adopted by the specification can achieve the following beneficial effects:
the model training method can be seen in that service data of online services within a set time can be collected, a service data set is constructed, then a preset classification model is trained according to the service data set and a sample data set used in training a wind control model, the classification model is used for judging which data set the input data belongs to, then the training quality degree of the trained classification model is determined, the integral discrimination of the service data set and the sample data set is determined according to the training quality degree, the wind control model is trained according to the discrimination, and wind control is performed on an online user through the trained wind control model.
As can be seen from the above, in the model training method provided in this specification, training needs to be performed through a service data set and a sample data set, and indirectly through the training quality of a trained classification model, the degree of distinction of the service data set and the sample data set as a whole is determined, that is, how much difference exists between sample distributions of the service data set and the sample data set is determined, so as to determine whether further iterative training needs to be performed on a wind control model, and how to train the wind control model, thereby improving the accuracy of wind control performed through the wind control model.
Drawings
The accompanying drawings, which are included to provide a further understanding of the specification and are incorporated in and constitute a part of this specification, illustrate embodiments of the specification and together with the description serve to explain the specification and not to limit the specification in a non-limiting sense. In the drawings:
FIG. 1 is a schematic flow chart of a method of model training in the present specification;
fig. 2 is a schematic flowchart of a process for determining a degree of distinction between a service data set and a sample data set by training a classification model provided in this specification;
FIG. 3 is a schematic diagram of a model training apparatus provided herein;
fig. 4 is a schematic diagram of an electronic device corresponding to fig. 1 provided in the present specification.
Detailed Description
In order to make the objects, technical solutions and advantages of the present disclosure more clear, the technical solutions of the present disclosure will be clearly and completely described below with reference to the specific embodiments of the present disclosure and the accompanying drawings. It is to be understood that the embodiments described are only a few embodiments of the present disclosure, and not all embodiments. All other embodiments obtained by a person of ordinary skill in the art based on the embodiments in the present specification without making any creative effort belong to the protection scope of the present specification.
The technical solutions provided by the embodiments of the present description are described in detail below with reference to the accompanying drawings.
Fig. 1 is a schematic flow chart of a model training method in this specification, which specifically includes the following steps:
s100: and acquiring service data of online services in set time, and constructing a service data set.
S102: and training a preset classification model according to the business data set and a sample data set used for training the wind control model, and training the preset classification model according to the business data set and the sample data set used for training the wind control model, wherein the classification model is used for judging which data set of the business data set and the sample data set the input data belongs to.
S104: and determining the training quality degree of the trained classification model.
In practical application, the risk faced by a user and a platform can be reduced through accurate wind control, due to the attack and defense confrontation attribute of a wind control scene, the model after online is often subjected to the condition of model performance degradation, the model degradation is caused by many reasons, the most direct reason is that the sample distribution encountered by the online environment is inconsistent with the sample distribution during training, and therefore the model training method provided by the specification is mainly used for avoiding the problem of performance degradation caused by the model after online along with the time lapse.
Specifically, the server may collect service data of an online service within a set time, construct a service data set, and train a preset classification model according to the service data set and a sample data set used when training the wind control model, where the classification model is used to determine to which data set of the service data set and the sample data set the input data belongs.
Then, the overall discrimination of the service data set and the sample data set can be determined according to the trained classification model. The wind control model mentioned here is a model for performing wind control on line, and in a service scenario where the wind control is performed through the wind control model is not limited, for example, the wind control model may be used for performing wind control on an account of a user on line to avoid fraudulent use of the account of the user, and for example, the wind control model may be used for performing wind control on a transaction on line to avoid a risky transaction.
The above-mentioned service data may be used for online controlling the transaction of the user, for example, the service data may be transaction data of the user, and the transaction data may be used for controlling the transaction of the user. The sample data set may refer to a training sample set (which may include all training samples for training the current wind control model) for training the current wind control model.
The classification model is a model for classifying between a business data set and a sample data set, that is, the training objective of the classification model mentioned herein is mainly to enable the classification model to distinguish the business data set from the sample data set, that is, for the classification model, the business data set is one category, and the sample data set is another category.
It should be noted that, the training quality degree of the trained classification model may be specifically determined, and the overall degree of distinction between the service data set and the sample data set is determined according to the training quality degree, where the training quality degree is used to represent whether the training effect on the classification model is good or not (in addition, the training quality degree may also be used to represent the classification effect on the classification model after training), where the higher the training quality degree is, the better the classification effect of the classification model may be, and a positive correlation may be formed between the classification quality degree and the degree of distinction. Of course, the degree of training quality can also be directly used as the discrimination.
Because if there is no difference in sample distribution between the service data set and the sample data set, it is difficult for the training model to classify the service data set and the sample data set well in the process of training the classification model, no matter how the classification model is trained, the classification effect of the classification model is poor, which can be called as, the quality of training the classification model by the service data set and the sample data set is low, so the degree of the quality of training is also low, therefore, the training quality degree can indirectly explain the overall discrimination between the service data set and the sample data set, therefore, if the training quality degree is lower, the discrimination is also lower, if the training quality degree is higher, the classification effect of the classification model is better, the classification model can distinguish the business data set from the sample data set, and the overall distinction degree between the business data set and the sample data set is higher.
Of course, the training quality degree calculated in a certain manner may be opposite to the training quality of the classification model, and if the training quality of the classification model is worse, the training quality degree is higher, and the degree of distinction between the service data set and the sample data set is lower as a whole.
The method for determining the training quality degree may be various, for example, a service data set and a sample data set may be input into a trained classification model to obtain a classification result; and determining the classification accuracy and the classification error rate corresponding to the trained classification model according to the classification result, determining a classification effect distribution curve corresponding to the trained classification model according to the classification accuracy and the classification error rate, and determining the training quality degree according to a Receiver Operating Characteristic (ROC) distribution curve.
Specifically, when the classification accuracy and the classification error rate are determined, the sample data set may be used as a positive sample set, the service data set may be used as a negative sample set, the positive sample set and the negative sample set are input into the trained classification model, the classification result for the positive sample set and the classification result for the negative sample set are obtained, the classification accuracy is obtained according to the proportion of correctly identified positive samples in all positive samples, and the classification error rate is obtained according to the proportion of incorrectly identified negative samples as positive samples in all negative samples.
The ROC curve, which is a curve passing through (0, 0) and (1,1) and located above the line connecting (0, 0) and (1,1), can be plotted with the classification accuracy as an ordinate and the classification error rate as an abscissa. The integral of the Curve, that is, the Area Under the Curve (AUC) of the Curve can represent the classification effect of the classification model, so that the integral corresponding to the Curve can be calculated as the training quality degree, and if the training quality degree is not less than the set value, the wind control model can be iteratively trained according to the service data set.
Certainly, the ROC curve may be drawn in a continuous process in a training process, that is, the process of training the classification model may be divided into a process of multiple times of training, and after each training, the classification accuracy and the classification error rate of the classification model may be determined through the service data set and part of data in the sample data set (as data of the verification set), so that the ROC curve is obtained through the classification accuracy and the classification error rate determined multiple times.
It should be noted that, in order to obtain the above-mentioned discrimination more accurately by training the classification model, the classification model may be constructed according to a model structure corresponding to the wind control model, where the model structures of the wind control model and the classification model may be identical, or the classification model may be constructed according to a specified network layer in the model structure corresponding to the wind control model, where the specified network layer may include: the method includes a feature extraction layer, where a data feature dimension of data input into the wind control model may be the same as a data feature dimension of data input into the classification model, an algorithm used by the classification model in this specification may be the same as the wind control model, and a method for processing feature data of the data by the classification model may be the same as the wind control model (this is also a reason for specifying that the network layer includes the feature extraction layer).
S106: and determining the integral discrimination of the business data set and the sample data set according to the training quality degree.
S108: and training the wind control model according to the discrimination so as to carry out wind control on the online user through the trained wind control model.
In the above, it is mentioned that the overall discrimination of the service data set and the sample data set can be determined according to the training quality degree, then, the overall discrimination of the service data set and the sample data set is obtained, and the wind control model can be trained according to the discrimination, so as to perform wind control on the online user through the trained wind control model.
The overall distinction degree of the service data set and the sample data set may refer to the distinction degree of the service data set and the sample data set on the sample distribution, when the sample distribution of the service data set and the sample data set has a certain distinction degree, it may be that data distribution under a certain characteristic dimension or certain characteristic dimensions between the service data set and the sample data set has a certain distinction, for a piece of service data or a piece of sample data, there are many characteristic dimensions, for example, if the service data is transaction data, the characteristic dimensions may include transaction time, transaction place, transaction object, transaction amount, and the like, and therefore, the characteristic dimensions referred to herein may refer to data of each dimension (or may be referred to as data of each category) included in the service data or the sample data.
Specifically, according to the above-mentioned discrimination, there are various ways to train the wind control model, and since the greater the discrimination between the service data set and the sample data set as a whole, the worse the effect of the wind control model on wind control may be, if it is determined that the discrimination is not less than the set discrimination, the wind control model may be trained according to the service data set. Of course, at least part of the data sets can be screened from the business data set according to the discrimination, the wind control model is trained, more data sets can be screened as the discrimination is higher, the wind control model is trained, and specifically, the proportion of the data sets to be screened from the business data set can be determined according to the discrimination, so that at least part of the data sets can be screened from the business data set, the wind control model is trained, and the higher the discrimination is, the higher the proportion can be.
In order to ensure that whether training data of the wind control model is different from business data at intervals is detected, so that whether the model effect of the wind control model is degraded or not is indirectly determined, online business data generated in a business period can be acquired according to each preset business period, the classification model is trained according to the online business data generated in the business period and the training data used for training the wind control model, the training quality degree of the trained classification model is obtained, the discrimination is determined according to the training quality degree of the trained classification model, and the wind control model is trained according to the discrimination, so that wind control is performed through the trained wind control model in the next business period.
For example, the duration of one service period may be set to 7 days, then a batch of service data may be collected every 7 days, and a classification model is trained through a service data set corresponding to the batch of service data and a sample data set of a wind control model, and a training quality degree of the trained classification model is determined.
In the above, the embodiments of the model training method in the present specification are described separately, and the method of the model training will be fully described below by way of an example, as shown in fig. 2.
Fig. 2 is a schematic flowchart of a process for determining a degree of distinction between a business data set and a sample data set by training a classification model provided in this specification.
As can be seen from fig. 2, in the process of training the classification model, the sample data set may be used as a positive sample, and the service data set may be used as a negative sample, that is, the classification model may be substantially a binary classification model, the sample data set is a class, the service data set is another class, the service data set is data generated on line, and therefore may be used for the on-line wind control of the wind control model, the sample data set is a training sample that has been historically trained on the wind control model, after the classification model is trained, the training quality degree of the classification model may be determined, the higher the training quality degree is, it indicates that the higher the overall degree of distinction between the service data set and the sample data set is, the lower the training quality degree is, it may indirectly indicate that the overall degree of distinction between the service data set and the sample data set is lower, and thus, according to the degree of distinction, for example, when the degree of distinction is higher than the set threshold, the wind control model may be trained by the service data set, wherein the service data set may be added to the sample data set, and the wind control model may be trained by the sample data set to which the service data set is added. Of course, according to the discrimination, part of the data set can be selected from the sample data set, and the wind control model can be trained.
Thus, it can be seen that the model training method provided in this specification can be implemented by using the newly generated business data on line and the training data for training the wind control model in the past, training the classification model, if the sample distribution of the business data is unchanged relative to the training data, the classification effect of the classification model after training is poor, if the sample distribution of the business data is changed to a certain extent relative to the training data, the classification effect after the classification model training is improved, therefore, the integral discrimination between the service data set and the sample data set can be indirectly obtained through the training quality degree capable of representing the classification effect of the classification model, if the discrimination is higher, the sample distribution of the online service data is changed to a certain extent, and the wind control model needs to be further iterated, so that the accuracy of wind control through the wind control model is improved.
The above method for model training provided for one or more embodiments of the present specification also provides a device for model training, as shown in fig. 3, based on the same idea.
Fig. 3 is a schematic diagram of a model training apparatus provided in this specification, which specifically includes:
the acquisition module 301 is configured to acquire service data of an online service within a set time and construct a service data set;
a training module 302, configured to train a preset classification model according to the service data set and a sample data set used for training the wind control model, where the classification model is used to determine to which data set of the service data set and the sample data set the input data belongs;
a first determining module 303, configured to determine a training quality degree of the trained classification model;
a second determining module 304, configured to determine, according to the training quality degree, a degree of distinction of the service data set and the sample data set as a whole;
and the wind control module 305 is configured to train the wind control model according to the discrimination so as to perform wind control on the online user through the trained wind control model.
Optionally, the first determining module 303 is specifically configured to input the service data set and the sample data set into the trained classification model to obtain a classification result; determining the classification accuracy and the classification error rate corresponding to the trained classification model according to the classification result; determining a classification effect distribution curve corresponding to the trained classification model according to the classification accuracy and the classification error rate; and determining the training quality degree according to the classification effect distribution curve.
Optionally, the first determining module 303 is specifically configured to use the sample data set as a positive sample set, use the service data set as a negative sample set, and input the positive sample set and the negative sample set into the trained classification model to obtain a classification result for the positive sample set and a classification result for the negative sample set; and obtaining a classification correct rate according to the proportion of the correctly identified positive samples in all the positive samples, and obtaining a classification error rate according to the proportion of the incorrectly identified negative samples as the positive samples in all the negative samples.
Optionally, the wind control module 305 is specifically configured to, if it is determined that the discrimination is not less than the set discrimination, train the wind control model according to the service data set.
Optionally, the apparatus further comprises:
the building module 306 is configured to build the classification model according to a model structure corresponding to the wind control model.
Optionally, the building module 306 is specifically configured to build the classification model according to a specified network layer in a model structure corresponding to the wind control model, where the specified network layer includes: and the characteristic extraction layer is used for inputting the data characteristic dimension of the data of the wind control model to be the same as the data characteristic dimension of the data of the classification model.
The present specification also provides a computer-readable storage medium having stored thereon a computer program, the computer program being operable to perform the method of model training described above.
This specification also provides a schematic block diagram of the electronic device shown in fig. 4. As shown in fig. 4, at the hardware level, the electronic device includes a processor, an internal bus, a network interface, a memory, and a non-volatile memory, and may also include hardware required for other services. The processor reads the corresponding computer program from the nonvolatile memory to the memory and then runs the computer program to realize the model training method. Of course, besides the software implementation, the present specification does not exclude other implementations, such as logic devices or a combination of software and hardware, and the like, that is, the execution subject of the following processing flow is not limited to each logic unit, and may be hardware or logic devices.
In the 90 s of the 20 th century, improvements in a technology could clearly distinguish between improvements in hardware (e.g., improvements in circuit structures such as diodes, transistors, switches, etc.) and improvements in software (improvements in process flow). However, as technology advances, many of today's process flow improvements have been seen as direct improvements in hardware circuit architecture. Designers almost always obtain a corresponding hardware circuit structure by programming an improved method flow into the hardware circuit. Thus, it cannot be said that an improvement in the process flow cannot be realized by hardware physical modules. For example, a Programmable Logic Device (PLD), such as a Field Programmable Gate Array (FPGA), is an integrated circuit whose Logic functions are determined by programming the Device by a user. A digital system is "integrated" on a PLD by the designer's own programming without requiring the chip manufacturer to design and fabricate application-specific integrated circuit chips. Furthermore, nowadays, instead of manually making an Integrated Circuit chip, such Programming is often implemented by "logic compiler" software, which is similar to a software compiler used in program development and writing, but the original code before compiling is also written by a specific Programming Language, which is called Hardware Description Language (HDL), and HDL is not only one but many, such as abel (advanced Boolean Expression Language), ahdl (alternate Hardware Description Language), traffic, pl (core universal Programming Language), HDCal (jhdware Description Language), lang, Lola, HDL, laspam, hardward Description Language (vhr Description Language), vhal (Hardware Description Language), and vhigh-Language, which are currently used in most common. It will also be apparent to those skilled in the art that hardware circuitry that implements the logical method flows can be readily obtained by merely slightly programming the method flows into an integrated circuit using the hardware description languages described above.
The controller may be implemented in any suitable manner, for example, the controller may take the form of, for example, a microprocessor or processor and a computer-readable medium storing computer-readable program code (e.g., software or firmware) executable by the (micro) processor, logic gates, switches, an Application Specific Integrated Circuit (ASIC), a programmable logic controller, and an embedded microcontroller, examples of which include, but are not limited to, the following microcontrollers: the ARC625D, Atmel AT91SAM, Microchip PIC18F26K20, and Silicone Labs C8051F320, the memory controller may also be implemented as part of the control logic for the memory. Those skilled in the art will also appreciate that, in addition to implementing the controller as pure computer readable program code, the same functionality can be implemented by logically programming method steps such that the controller is in the form of logic gates, switches, application specific integrated circuits, programmable logic controllers, embedded microcontrollers and the like. Such a controller may thus be considered a hardware component, and the means included therein for performing the various functions may also be considered as a structure within the hardware component. Or even means for performing the functions may be regarded as being both a software module for performing the method and a structure within a hardware component.
The systems, devices, modules or units illustrated in the above embodiments may be implemented by a computer chip or an entity, or by a product with certain functions. One typical implementation device is a computer. In particular, the computer may be, for example, a personal computer, a laptop computer, a cellular telephone, a camera phone, a smartphone, a personal digital assistant, a media player, a navigation device, an email device, a game console, a tablet computer, a wearable device, or a combination of any of these devices.
For convenience of description, the above devices are described as being divided into various units by function, and are described separately. Of course, the functions of the various elements may be implemented in the same one or more software and/or hardware implementations of the present description.
As will be appreciated by one skilled in the art, embodiments of the present invention may be provided as a method, system, or computer program product. Accordingly, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.
The present invention is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each flow and/or block of the flow diagrams and/or block diagrams, and combinations of flows and/or blocks in the flow diagrams and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
In a typical configuration, a computing device includes one or more processors (CPUs), input/output interfaces, network interfaces, and memory.
The memory may include forms of volatile memory in a computer readable medium, Random Access Memory (RAM) and/or non-volatile memory, such as Read Only Memory (ROM) or flash memory (flash RAM). Memory is an example of a computer-readable medium.
Computer-readable media, including both non-transitory and non-transitory, removable and non-removable media, may implement information storage by any method or technology. The information may be computer readable instructions, data structures, modules of a program, or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), Static Random Access Memory (SRAM), Dynamic Random Access Memory (DRAM), other types of Random Access Memory (RAM), Read Only Memory (ROM), Electrically Erasable Programmable Read Only Memory (EEPROM), flash memory or other memory technology, compact disc read only memory (CD-ROM), Digital Versatile Discs (DVD) or other optical storage, magnetic cassettes, magnetic tape magnetic disk storage or other magnetic storage devices, or any other non-transmission medium that can be used to store information that can be accessed by a computing device. As defined herein, a computer readable medium does not include a transitory computer readable medium such as a modulated data signal and a carrier wave.
It should also be noted that the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising an … …" does not exclude the presence of other like elements in a process, method, article, or apparatus that comprises the element.
As will be appreciated by one skilled in the art, embodiments of the present description may be provided as a method, system, or computer program product. Accordingly, the description may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the description may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.
This description may be described in the general context of computer-executable instructions, such as program modules, being executed by a computer. Generally, program modules include routines, programs, objects, components, data structures, etc. that perform particular tasks or implement particular abstract data types. The specification may also be practiced in distributed computing environments where tasks are performed by remote processing nodes that are linked through a communications network. In a distributed computing environment, program modules may be located in both local and remote computer storage media including memory nodes.
The embodiments in the present specification are described in a progressive manner, and the same and similar parts among the embodiments are referred to each other, and each embodiment focuses on the differences from the other embodiments. In particular, for the system embodiment, since it is substantially similar to the method embodiment, the description is simple, and for the relevant points, reference may be made to the partial description of the method embodiment.
The above description is only an example of the present specification, and is not intended to limit the present specification. Various modifications and alterations to this description will become apparent to those skilled in the art. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present specification should be included in the scope of the claims of the present specification.
Claims (10)
1. A method of model training, comprising:
acquiring service data of online service in set time, and constructing a service data set;
training a preset classification model according to the business data set and a sample data set used in training a wind control model, wherein the classification model is used for judging which data set of the business data set and the sample data set the input data belongs to;
determining the training quality degree of the trained classification model;
determining the integral discrimination of the business data set and the sample data set according to the training quality degree;
and training the wind control model according to the discrimination so as to perform wind control on the online user through the trained wind control model.
2. The method of claim 1, determining a degree of training quality of the trained classification model, comprising:
inputting the service data set and the sample data set into the trained classification model to obtain a classification result;
determining the classification accuracy and the classification error rate corresponding to the trained classification model according to the classification result;
determining a classification effect distribution curve corresponding to the trained classification model according to the classification accuracy and the classification error rate;
and determining the training quality degree according to the classification effect distribution curve.
3. The method of claim 2, inputting the traffic data set and the sample data set into the trained classification model to obtain a classification result, comprising:
taking the sample data set as a positive sample set and the service data set as a negative sample set, and inputting the positive sample set and the negative sample set into the trained classification model to obtain a classification result aiming at the positive sample set and a classification result aiming at the negative sample set;
according to the classification result, determining the classification accuracy and the classification error rate corresponding to the trained classification model, specifically comprising:
and obtaining the classification correct rate according to the proportion of the correctly identified positive samples in all the positive samples, and obtaining the classification error rate according to the proportion of the incorrectly identified negative samples as the positive samples in all the negative samples.
4. The method of claim 1, training the wind control model based on the discrimination, comprising:
and if the discrimination is not smaller than the set discrimination, training the wind control model according to the service data set.
5. The method of claim 1, wherein before training the preset classification model according to the traffic data set and the training data set, the method further comprises:
and constructing the classification model according to the model structure corresponding to the wind control model.
6. The method of claim 1, wherein constructing the classification model according to the model structure corresponding to the wind control model comprises:
constructing the classification model according to an appointed network layer in a model structure corresponding to the wind control model, wherein the appointed network layer comprises: and the characteristic extraction layer is used for inputting the data characteristic dimension of the data of the wind control model to be the same as the data characteristic dimension of the data of the classification model.
7. An apparatus for model training, comprising:
the acquisition module is used for acquiring service data of online services in set time and constructing a service data set;
the training module is used for training a preset classification model according to the business data set and a sample data set used in training the wind control model, and the classification model is used for judging which data set of the business data set and the sample data set the input data belongs to;
the first determining module is used for determining the training quality degree of the trained classification model;
the second determining module is used for determining the integral discrimination of the business data set and the sample data set according to the training quality degree;
and the wind control module is used for training the wind control model according to the discrimination so as to perform wind control on the online user through the trained wind control model.
8. The apparatus according to claim 7, wherein the first determining module is specifically configured to input the service data set and the sample data set into the trained classification model to obtain a classification result; determining the classification accuracy and the classification error rate corresponding to the trained classification model according to the classification result; determining a classification effect distribution curve corresponding to the trained classification model according to the classification accuracy and the classification error rate; and determining the training quality degree according to the classification effect distribution curve.
9. A computer-readable storage medium, in which a computer program is stored which, when being executed by a processor, carries out the method of any one of the preceding claims 1 to 6.
10. An electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, the processor implementing the method of any of the preceding claims 1 to 6 when executing the program.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210752076.4A CN114943307A (en) | 2022-06-28 | 2022-06-28 | Model training method and device, storage medium and electronic equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210752076.4A CN114943307A (en) | 2022-06-28 | 2022-06-28 | Model training method and device, storage medium and electronic equipment |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114943307A true CN114943307A (en) | 2022-08-26 |
Family
ID=82911530
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210752076.4A Pending CN114943307A (en) | 2022-06-28 | 2022-06-28 | Model training method and device, storage medium and electronic equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114943307A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115543945A (en) * | 2022-11-29 | 2022-12-30 | 支付宝(杭州)信息技术有限公司 | Model compression method and device, storage medium and electronic equipment |
| CN115545353A (en) * | 2022-11-29 | 2022-12-30 | 支付宝(杭州)信息技术有限公司 | Method and device for business wind control, storage medium and electronic equipment |
| CN115563584A (en) * | 2022-11-29 | 2023-01-03 | 支付宝(杭州)信息技术有限公司 | Model training method and device, storage medium and electronic equipment |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109598281A (en) * | 2018-10-11 | 2019-04-09 | 阿里巴巴集团控股有限公司 | A kind of business risk preventing control method, device and equipment |
| US20190164062A1 (en) * | 2017-11-28 | 2019-05-30 | International Business Machines Corporation | Data classifier |
| WO2019114412A1 (en) * | 2017-12-15 | 2019-06-20 | 阿里巴巴集团控股有限公司 | Graphical structure model-based method for credit risk control, and device and equipment |
| US20200211106A1 (en) * | 2017-12-27 | 2020-07-02 | Alibaba Group Holding Limited | Method, apparatus, and device for training risk management models |
| CN113344700A (en) * | 2021-07-27 | 2021-09-03 | 上海华瑞银行股份有限公司 | Wind control model construction method and device based on multi-objective optimization and electronic equipment |
| CN113643119A (en) * | 2021-07-19 | 2021-11-12 | 北京三快在线科技有限公司 | Model training method, business wind control method and business wind control device |
| CN114078050A (en) * | 2021-11-17 | 2022-02-22 | 中国建设银行股份有限公司 | Loan overdue prediction method and device, electronic equipment and computer readable medium |
-
2022
- 2022-06-28 CN CN202210752076.4A patent/CN114943307A/en active Pending
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20190164062A1 (en) * | 2017-11-28 | 2019-05-30 | International Business Machines Corporation | Data classifier |
| WO2019114412A1 (en) * | 2017-12-15 | 2019-06-20 | 阿里巴巴集团控股有限公司 | Graphical structure model-based method for credit risk control, and device and equipment |
| US20200211106A1 (en) * | 2017-12-27 | 2020-07-02 | Alibaba Group Holding Limited | Method, apparatus, and device for training risk management models |
| CN109598281A (en) * | 2018-10-11 | 2019-04-09 | 阿里巴巴集团控股有限公司 | A kind of business risk preventing control method, device and equipment |
| CN113643119A (en) * | 2021-07-19 | 2021-11-12 | 北京三快在线科技有限公司 | Model training method, business wind control method and business wind control device |
| CN113344700A (en) * | 2021-07-27 | 2021-09-03 | 上海华瑞银行股份有限公司 | Wind control model construction method and device based on multi-objective optimization and electronic equipment |
| CN114078050A (en) * | 2021-11-17 | 2022-02-22 | 中国建设银行股份有限公司 | Loan overdue prediction method and device, electronic equipment and computer readable medium |
Non-Patent Citations (1)
| Title |
|---|
| 陈梦珲: "基于卷积神经网络的金融风控模型设计与实现", 万方, 27 August 2019 (2019-08-27) * |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115543945A (en) * | 2022-11-29 | 2022-12-30 | 支付宝(杭州)信息技术有限公司 | Model compression method and device, storage medium and electronic equipment |
| CN115545353A (en) * | 2022-11-29 | 2022-12-30 | 支付宝(杭州)信息技术有限公司 | Method and device for business wind control, storage medium and electronic equipment |
| CN115563584A (en) * | 2022-11-29 | 2023-01-03 | 支付宝(杭州)信息技术有限公司 | Model training method and device, storage medium and electronic equipment |
| CN115563584B (en) * | 2022-11-29 | 2023-04-07 | 支付宝(杭州)信息技术有限公司 | Model training method and device, storage medium and electronic equipment |
| CN115545353B (en) * | 2022-11-29 | 2023-04-18 | 支付宝(杭州)信息技术有限公司 | Business wind control method, device, storage medium and electronic equipment |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108460523B (en) | Wind control rule generation method and device | |
| CN109544166B (en) | Risk identification method and risk identification device | |
| CN114943307A (en) | Model training method and device, storage medium and electronic equipment | |
| CN108596410B (en) | Automatic wind control event processing method and device | |
| CN111080304A (en) | Credible relationship identification method, device and equipment | |
| CN110008991B (en) | Risk event identification method, risk identification model generation method, risk event identification device, risk identification equipment and risk identification medium | |
| CN115618748B (en) | Model optimization method, device, equipment and storage medium | |
| CN115618964A (en) | Model training method and device, storage medium and electronic equipment | |
| CN110033092B (en) | Data label generation method, data label training device, event recognition method and event recognition device | |
| CN115035347A (en) | Picture identification method and device and electronic equipment | |
| CN113010562B (en) | Information recommendation method and device | |
| CN117313141A (en) | Abnormality detection method, abnormality detection device, abnormality detection equipment and readable storage medium | |
| CN112597255A (en) | Method and device for determining abnormal data | |
| CN110738562B (en) | Method, device and equipment for generating risk reminding information | |
| CN115567371B (en) | Abnormity detection method, device, equipment and readable storage medium | |
| CN116822606A (en) | Training method, device, equipment and storage medium of anomaly detection model | |
| CN111538925A (en) | Method and device for extracting Uniform Resource Locator (URL) fingerprint features | |
| CN114511376A (en) | Credit data processing method and device based on multiple models | |
| CN109325127B (en) | Risk identification method and device | |
| CN113344590A (en) | Method and device for model training and complaint rate estimation | |
| CN111242195A (en) | Model, insurance wind control model training method and device and electronic equipment | |
| CN111461892A (en) | Method and device for selecting derived variables of risk identification model | |
| CN115564450B (en) | Wind control method, device, storage medium and equipment | |
| CN115545938B (en) | Method, device, storage medium and equipment for executing risk identification service | |
| CN112817948B (en) | Data detection method and device, readable storage medium 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 |