CN112766490A - Characteristic variable learning method, device, equipment and computer readable storage medium - Google Patents
Characteristic variable learning method, device, equipment and computer readable storage medium Download PDFInfo
- Publication number
- CN112766490A CN112766490A CN202110045805.8A CN202110045805A CN112766490A CN 112766490 A CN112766490 A CN 112766490A CN 202110045805 A CN202110045805 A CN 202110045805A CN 112766490 A CN112766490 A CN 112766490A
- Authority
- CN
- China
- Prior art keywords
- modeling
- learning
- variable
- characteristic variable
- different
- 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.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 61
- 230000006870 function Effects 0.000 claims abstract description 52
- 238000003062 neural network model Methods 0.000 claims abstract description 23
- 230000003993 interaction Effects 0.000 claims description 30
- 238000000586 desensitisation Methods 0.000 claims description 20
- 238000005457 optimization Methods 0.000 claims description 13
- 238000013507 mapping Methods 0.000 claims description 9
- 238000005070 sampling Methods 0.000 claims description 9
- 230000000694 effects Effects 0.000 abstract description 6
- 238000005516 engineering process Methods 0.000 abstract description 6
- 230000008569 process Effects 0.000 description 7
- 238000004891 communication Methods 0.000 description 6
- 238000010586 diagram Methods 0.000 description 4
- 238000012549 training Methods 0.000 description 4
- 238000004364 calculation method Methods 0.000 description 2
- 210000002569 neuron Anatomy 0.000 description 2
- 238000012545 processing Methods 0.000 description 2
- 238000013526 transfer learning Methods 0.000 description 2
- 238000013473 artificial intelligence Methods 0.000 description 1
- 230000006399 behavior Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 238000013178 mathematical model Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000012954 risk control Methods 0.000 description 1
- 238000012216 screening Methods 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
- 238000012546 transfer Methods 0.000 description 1
Images
Classifications
-
- 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
-
- 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
- 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
- G06Q40/03—Credit; Loans; Processing thereof
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- General Health & Medical Sciences (AREA)
- Accounting & Taxation (AREA)
- Software Systems (AREA)
- Mathematical Physics (AREA)
- General Engineering & Computer Science (AREA)
- Computing Systems (AREA)
- Molecular Biology (AREA)
- Finance (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- Marketing (AREA)
- Development Economics (AREA)
- General Business, Economics & Management (AREA)
- Economics (AREA)
- Technology Law (AREA)
- Strategic Management (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
- Machine Translation (AREA)
Abstract
The invention relates to the technical field of financial technology (Fintech). The invention discloses a characteristic variable learning method, a device, equipment and a computer readable storage medium, wherein modeling characteristic variable sets corresponding to different modeling tasks of different types of customer groups are obtained and are used as the input of the same target model, so that the different customer groups and the different modeling tasks do not need to be modeled respectively; and then, a target model is built based on a neural network model and a multi-task learning algorithm, a uniform input variable expression, namely a final target function, is learned for different modeling tasks in a mode of sharing hidden layer parameters, so that the optimal effect is achieved on different modeling tasks as far as possible, model parameters are prevented from being excessively fitted to a certain biased customer group distribution, and the conditions that the model is difficult to maintain, modeling intermediate information is recycled or the model cannot be built and the like caused by respective modeling in the existing mode are avoided.
Description
Technical Field
The invention relates to the technical field of financial technology (Fintech), in particular to a characteristic variable learning method, a device, equipment and a computer readable storage medium.
Background
With the development of computer technology, more and more technologies (big data, distributed, Blockchain, artificial intelligence, etc.) are applied to the financial field, and the traditional financial industry is gradually changing to financial technology (Fintech), but higher requirements are also put forward on the technologies due to the requirements of security and real-time performance of the financial industry. For the credit risk modeling field, because customers often have various different stages in credit business, such as account opening, first loan, inventory, overdue, and the like, the existing modeling manner is to establish multiple models according to different customer groups in different business stages, but the multiple models established based on the method are difficult to maintain due to large number, and even the multiple models may not be successfully established in some cases. Therefore, the above situations all reflect the technical problem that the existing credit risk modeling mode is difficult to realize.
Disclosure of Invention
The invention mainly aims to provide a characteristic variable learning method, a characteristic variable learning device, characteristic variable learning equipment and a computer readable storage medium, and aims to solve the technical problem that the existing credit risk modeling mode is difficult to realize.
In order to achieve the above object, the present invention provides a feature variable learning method, including:
the method comprises the steps of obtaining modeling characteristic variable sets corresponding to modeling sample clients in different business stages, wherein the modeling sample clients are obtained based on different client groups, and the modeling characteristic variable sets correspond to different modeling tasks;
building a target model based on a neural network model and a multi-task learning algorithm framework, and combining the modeling characteristic variable set as the input of the target model so as to perform parameter sharing on the modeling characteristic variable set in the target model to obtain a parameter sharing result;
and performing parallel learning on the parameter sharing result aiming at the modeling tasks to obtain an objective function through learning, wherein the objective function is used for uniformly representing the mapping relation between the modeling characteristic variable set and the modeling tasks.
Optionally, the building a target model based on a neural network model and a multi-task learning algorithm architecture, and the step of using the modeling feature variable set as an input of the target model to perform parameter sharing on the modeling feature variable set in the target model to obtain a parameter sharing result includes:
building a model comprising an input variable layer, a shared hidden layer, a shared output layer and a task output layer by using a neural network model architecture and a multi-task learning algorithm to serve as the target model;
and inputting the modeling characteristic variable set into an input variable layer of the target model, so as to transmit the modeling characteristic variable set from the input variable layer into the shared hidden layer, and performing deep interaction between variables in the shared hidden layer to obtain the parameter sharing result.
Optionally, the step of transferring the modeling feature variable set from the input variable layer to the shared hidden layer, and performing deep interaction between variables in the shared hidden layer to obtain the parameter sharing result includes:
transmitting the modeling characteristic variable set from the input variable layer to the shared hidden layer so that the modeling characteristic variable set learns multiple groups of universal variable interaction information in the shared hidden layer;
and transmitting the multiple groups of the general variable interaction information into the shared output layer so as to summarize the multiple groups of the general variable interaction information into the parameter sharing result on the shared output layer.
Optionally, the step of learning the parameter sharing result in parallel for a plurality of modeling tasks to obtain an objective function includes:
transmitting the parameter sharing result into the task output layer, and learning to obtain a gradient optimization direction based on a plurality of modeling tasks on the task output layer;
and optimizing the parameter sharing result by utilizing the gradient optimization direction to obtain the objective function based on the optimized parameter sharing result.
Optionally, the step of obtaining modeling feature variable sets corresponding to modeling sample clients in a plurality of different business phases includes:
acquiring customer information and business stage information corresponding to the customer information, and dividing the customer information into a plurality of different customer groups according to the business stage information;
and sampling the modeling sample clients from a plurality of different client groups, and acquiring a modeling characteristic variable set corresponding to the modeling sample clients.
Optionally, the step of obtaining the customer information and the service phase information corresponding to the customer information, and dividing the customer information into a plurality of different customer groups according to the service phase information includes:
obtaining the client information and the service phase information, and carrying out desensitization treatment on the client information to obtain desensitization client information;
and carrying out guest group division on the desensitization guest information by using the service phase information to obtain a plurality of different guest groups.
Optionally, after the step of performing parallel learning on the parameter sharing result for a plurality of modeling tasks to obtain an objective function, the method further includes:
and determining a modeling task to be predicted in the modeling tasks, predicting and outputting a target output result of the modeling task to be predicted based on the target function.
Further, to achieve the above object, the present invention provides a characteristic variable learning device including:
the characteristic variable acquisition module is used for acquiring modeling characteristic variable sets corresponding to modeling sample clients in different business stages, wherein the modeling sample clients are obtained based on different client groups, and the modeling characteristic variable sets correspond to different modeling tasks;
the model parameter sharing module is used for building a target model based on a neural network model and a multi-task learning algorithm framework, and the modeling characteristic variable set is used as the input of the target model so as to perform parameter sharing on the modeling characteristic variable set in the target model to obtain a parameter sharing result;
and the target function learning module is used for performing parallel learning on the parameter sharing result aiming at the modeling tasks to obtain a target function through learning, wherein the target function is used for uniformly representing the mapping relation between the modeling characteristic variable set and the modeling tasks.
Optionally, the model parameter sharing module includes:
the target model building unit is used for building a model comprising an input variable layer, a shared hidden layer, a shared output layer and a task output layer as the target model by utilizing a neural network model architecture and a multi-task learning algorithm;
and the variable depth interaction unit is used for inputting the modeling characteristic variable set into an input variable layer of the target model, so that the modeling characteristic variable set is transmitted into the shared hidden layer from the input variable layer, and deep interaction among variables is performed in the shared hidden layer to obtain the parameter sharing result.
Optionally, the variable depth interaction unit includes:
transmitting the modeling characteristic variable set from the input variable layer to the shared hidden layer so that the modeling characteristic variable set learns multiple groups of universal variable interaction information in the shared hidden layer;
and transmitting the multiple groups of the general variable interaction information into the shared output layer so as to summarize the multiple groups of the general variable interaction information into the parameter sharing result on the shared output layer.
Optionally, the objective function learning module includes:
the optimization direction learning unit is used for transmitting the parameter sharing result into the task output layer, and learning on the basis of a plurality of modeling tasks in the task output layer to obtain a gradient optimization direction;
and the objective function obtaining unit is used for optimizing the parameter sharing result by utilizing the gradient optimization direction so as to obtain the objective function based on the optimized parameter sharing result.
Optionally, the feature variable obtaining module includes:
the service guest group dividing unit is used for acquiring guest information and service stage information corresponding to the guest information, and dividing the guest group of the guest information according to the service stage information to obtain a plurality of different guest groups;
and the characteristic variable acquisition unit is used for sampling the modeling sample clients from a plurality of different client groups and acquiring the modeling characteristic variable sets corresponding to the modeling sample clients.
Optionally, the service guest group dividing unit includes:
a client information desensitization unit, configured to obtain the client information and the service phase information, and perform desensitization processing on the client information to obtain desensitized client information;
and the desensitization information dividing unit is used for carrying out guest group division on the desensitization customer information by using the service phase information to obtain a plurality of different guest groups.
Optionally, the feature variable learning apparatus further includes:
and the target result output unit is used for determining a modeling task to be predicted in a plurality of modeling tasks, predicting and outputting a target output result of the modeling task to be predicted based on the target function.
Further, to achieve the above object, the present invention also provides a feature variable learning apparatus including: a memory, a processor and a feature variable learning program stored on the memory and executable on the processor, the feature variable learning program when executed by the processor implementing the steps of the feature variable learning method as described above.
Further, to achieve the above object, the present invention also provides a computer-readable storage medium having stored thereon a feature variable learning program which, when executed by a processor, implements the steps of the feature variable learning method as described above.
The invention provides a characteristic variable learning method, a characteristic variable learning device, characteristic variable learning equipment and a computer readable storage medium. According to the method, the modeling characteristic variable sets corresponding to different modeling tasks of different types of customer groups are obtained and serve as the input of the same target model, so that the different customer groups and the different modeling tasks do not need to be modeled respectively; then, a target model is built based on a neural network model and a multi-task learning algorithm, a uniform input variable expression, namely a final target function, is learned for different modeling tasks in a mode of sharing hidden layer parameters, so that the optimal effect is achieved on the different modeling tasks as far as possible, model parameters are prevented from being excessively fitted to a certain biased passenger group distribution, the situations that the model is difficult to maintain, modeling intermediate information is recycled or the model cannot be built and the like caused by respective modeling in the existing mode are avoided, and the technical problem that the realization difficulty of the existing credit risk modeling mode is large is solved.
Drawings
FIG. 1 is a schematic diagram of an apparatus architecture of a hardware operating environment according to an embodiment of the present invention;
FIG. 2 is a flowchart illustrating a characteristic variable learning method according to a first embodiment of the present invention;
FIG. 3 is a schematic diagram of a model principle of a second embodiment of the characteristic variable learning method of the present invention;
FIG. 4 is a functional block diagram of the characteristic variable learning apparatus according to the present invention.
The implementation, functional features and advantages of the objects of the present invention will be further explained with reference to the accompanying drawings.
Detailed Description
It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.
As shown in fig. 1, fig. 1 is a schematic device structure diagram of a hardware operating environment according to an embodiment of the present invention.
As shown in fig. 1, the characteristic variable learning means may include: a processor 1001, such as a CPU, a user interface 1003, a network interface 1004, a memory 1005, a communication bus 1002. Wherein a communication bus 1002 is used to enable connective communication between these components. The user interface 1003 may include a Display screen (Display), an input unit such as a Keyboard (Keyboard), and the optional user interface 1003 may also include a standard wired interface, a wireless interface. The network interface 1004 may optionally include a standard wired interface, a wireless interface (e.g., WI-FI interface). The memory 1005 may be a high-speed RAM memory or a non-volatile memory (e.g., a magnetic disk memory). The memory 1005 may alternatively be a storage device separate from the processor 1001.
Those skilled in the art will appreciate that the configuration of the apparatus shown in fig. 1 is not intended to be limiting of the apparatus and may include more or fewer components than those shown, or some components may be combined, or a different arrangement of components.
As shown in fig. 1, a memory 1005, which is a kind of computer storage medium, may include therein an operating system, a network communication module, a user interface module, and a feature variable learning program.
In the device shown in fig. 1, the network interface 1004 is mainly used for connecting to a backend server and performing data communication with the backend server; the user interface 1003 is mainly used for connecting a client (programmer's end) and performing data communication with the client; and the processor 1001 may be configured to call the feature variable learning program stored in the memory 1005, and perform the following operations in the feature variable learning method:
the method comprises the steps of obtaining modeling characteristic variable sets corresponding to modeling sample clients in different business stages, wherein the modeling sample clients are obtained based on different client groups, and the modeling characteristic variable sets correspond to different modeling tasks;
building a target model based on a neural network model and a multi-task learning algorithm framework, and combining the modeling characteristic variable set as the input of the target model so as to perform parameter sharing on the modeling characteristic variable set in the target model to obtain a parameter sharing result;
and performing parallel learning on the parameter sharing result aiming at the modeling tasks to obtain an objective function through learning, wherein the objective function is used for uniformly representing the mapping relation between the modeling characteristic variable set and the modeling tasks.
Further, the step of building a target model based on a neural network model and a multitask learning algorithm architecture, and using the modeling characteristic variable set as an input of the target model to perform parameter sharing on the modeling characteristic variable set in the target model to obtain a parameter sharing result includes:
building a model comprising an input variable layer, a shared hidden layer, a shared output layer and a task output layer by using a neural network model architecture and a multi-task learning algorithm to serve as the target model;
and inputting the modeling characteristic variable set into an input variable layer of the target model, so as to transmit the modeling characteristic variable set from the input variable layer into the shared hidden layer, and performing deep interaction between variables in the shared hidden layer to obtain the parameter sharing result.
Further, the step of transferring the modeling feature variable set from the input variable layer to the shared hidden layer, and performing deep interaction between variables in the shared hidden layer to obtain the parameter sharing result includes:
transmitting the modeling characteristic variable set from the input variable layer to the shared hidden layer so that the modeling characteristic variable set learns multiple groups of universal variable interaction information in the shared hidden layer;
and transmitting the multiple groups of the general variable interaction information into the shared output layer so as to summarize the multiple groups of the general variable interaction information into the parameter sharing result on the shared output layer.
Further, the step of learning the parameter sharing result in parallel for a plurality of modeling tasks to obtain an objective function includes:
transmitting the parameter sharing result into the task output layer, and learning to obtain a gradient optimization direction based on a plurality of modeling tasks on the task output layer;
and optimizing the parameter sharing result by utilizing the gradient optimization direction to obtain the objective function based on the optimized parameter sharing result.
Further, the step of obtaining modeling feature variable sets corresponding to modeling sample clients in a plurality of different business phases includes:
acquiring customer information and business stage information corresponding to the customer information, and dividing the customer information into a plurality of different customer groups according to the business stage information;
and sampling the modeling sample clients from a plurality of different client groups, and acquiring a modeling characteristic variable set corresponding to the modeling sample clients.
Further, the step of obtaining the customer information and the service phase information corresponding to the customer information, and dividing the customer information into a plurality of different customer groups according to the service phase information includes:
obtaining the client information and the service phase information, and carrying out desensitization treatment on the client information to obtain desensitization client information;
and carrying out guest group division on the desensitization guest information by using the service phase information to obtain a plurality of different guest groups.
Further, after the step of learning the parameter sharing result in parallel for several modeling tasks to obtain the target function, the processor 1001 may be configured to call a feature variable learning program stored in the memory 1005 and perform the following operations in the feature variable learning method:
and determining a modeling task to be predicted in the modeling tasks, predicting and outputting a target output result of the modeling task to be predicted based on the target function.
Based on the hardware structure, the embodiment of the characteristic variable learning method is provided.
In order to solve the problems, the invention provides a characteristic variable learning method, namely, modeling characteristic variable sets corresponding to different modeling tasks of different types of customer groups are obtained and are used as the input of the same target model, so that different customer groups and different modeling tasks do not need to be modeled respectively; then, a target model is built based on a neural network model and a multi-task learning algorithm, a uniform input variable expression, namely a final target function, is learned for different modeling tasks in a mode of sharing hidden layer parameters, so that the optimal effect is achieved on the different modeling tasks as far as possible, model parameters are prevented from being excessively fitted to a certain biased passenger group distribution, the situations that the model is difficult to maintain, modeling intermediate information is recycled or the model cannot be built and the like caused by respective modeling in the existing mode are avoided, and the technical problem that the realization difficulty of the existing credit risk modeling mode is large is solved.
Referring to fig. 2, fig. 2 is a flowchart illustrating a characteristic variable learning method according to a first embodiment of the present invention. The characteristic variable learning method comprises the following steps;
step S10, obtaining modeling characteristic variable sets corresponding to modeling sample clients in different business stages, wherein the modeling sample clients are obtained based on different client groups, and the modeling characteristic variable sets correspond to different modeling tasks;
in the present embodiment, the present invention is applied to a terminal device. The method is mainly applicable to application scenarios of credit risk modeling. For personal credit business, credit risk modeling is to identify the personal risk of default by combining various factors causing the personal default and utilizing a mathematical model method, and the personal risk of default is used in the whole risk control process. The business stage refers to different stages of the credit business of the client, such as account opening, first borrowing, stock, overdue and the like, and different client groups can be divided at different stages to correspond to different risk modeling tasks. The modeling sample client refers to a client object for acquiring modeling characteristic variables at this time. The modeling feature variable set comprises a plurality of modeling feature variables, and the modeling feature variables refer to behavior information of modeling sample customers in front of observation points and are represented in a digital form, and are generally represented by x. The observation point refers to the point at which the customer begins to present the risk. The customer group refers to a customer group which is screened by certain screening conditions and has a plurality of common characteristics. The modeling task refers to whether the client becomes a target client in the expression period of the client in different business phases, and is generally represented by y, and the specific value can be represented by marking the target client as 1 and marking the non-target client as 0.
Specifically, if the terminal receives a characteristic variable learning instruction, a modeling sample client which needs to acquire a modeling characteristic variable of the terminal at present is determined according to the characteristic variable learning instruction, then modeling characteristic variables and the like of the modeling sample client within a certain time period are acquired from specified relevant platforms, and the data are collected into the modeling characteristic variable set.
Step S20, a target model is built based on a neural network model and a multi-task learning algorithm framework, the modeling characteristic variable set is used as the input of the target model, and the modeling characteristic variable set is subjected to parameter sharing in the target model to obtain a parameter sharing result;
in this embodiment, the neural network model refers to a machine learning algorithm, and the model includes an input layer, a hidden layer, and an output layer, each layer includes a plurality of neurons, the neurons between adjacent layers are connected by weighted edges, the forward calculation is a prediction process, and the backward calculation is a training process. The multi-task learning algorithm is one of transfer learning algorithms, and transfer learning can be understood as defining a source field and a target field, learning in the source field, and transferring learned knowledge to the target field, so that the learning effect of the target field is improved. The main goal of multi-task learning is to improve generalization ability by exploiting domain-specific information implicit in the training signals of multiple related tasks, which is accomplished by training multiple tasks in parallel using a shared representation. The parameter sharing result refers to feature variables of different customer groups and variable interaction information and parameters common to different modeling tasks, which are obtained by learning in a hidden layer of the model.
And the terminal establishes a neural network model sharing parameters, namely the target model, then uniformly takes all modeling characteristic variables in the modeling characteristic variable set as the input of the target model, carries out deep interaction between model parameters in the target model by utilizing a multi-target learning algorithm, and learns the model parameters by depending on the gradient returned from the upper layer to obtain the parameter sharing result.
Step S30, performing parallel learning on the parameter sharing result for the modeling tasks to obtain an objective function, where the objective function is used to uniformly represent a mapping relationship between the modeling feature variable set and the modeling tasks.
In this embodiment, the objective function is used to uniformly characterize the mapping relationship between the modeling feature variable set and the modeling tasks, specifically, refer to a uniform feature variable x expression for a plurality of different modeling tasks y.
And performing parameter learning sharing on the characteristic variables corresponding to the different modeling tasks in a hidden layer of the target model to obtain parameter sharing results, then further learning aiming at the different modeling tasks, obtaining a loss function finally obtained by weighting and summing target loss functions corresponding to the different modeling tasks after learning is completed, and finally obtaining the target function based on the loss function. In the prediction stage after learning is completed, one or more modeling tasks needing prediction at present can be selected to be accessed into the target model, and the model can output the prediction results of the modeling tasks needing prediction.
The invention provides a characteristic variable learning method. The characteristic variable learning method comprises the steps of obtaining modeling characteristic variable sets corresponding to modeling sample clients in a plurality of different service stages, wherein the modeling sample clients are obtained based on a plurality of different client groups, and the modeling characteristic variable sets correspond to a plurality of different modeling tasks; building a target model based on a neural network model and a multi-task learning algorithm framework, and combining the modeling characteristic variable set as the input of the target model so as to perform parameter sharing on the modeling characteristic variable set in the target model to obtain a parameter sharing result; and performing parallel learning on the parameter sharing result aiming at the modeling tasks to obtain an objective function through learning, wherein the objective function is used for uniformly representing the mapping relation between the modeling characteristic variable set and the modeling tasks. By the mode, the modeling feature variable sets corresponding to different modeling tasks of different types of customer groups are obtained and serve as the input of the same target model, so that different customer groups and different modeling tasks do not need to be modeled respectively; then, a target model is built based on a neural network model and a multi-task learning algorithm, a uniform input variable expression, namely a final target function, is learned for different modeling tasks in a mode of sharing hidden layer parameters, so that the optimal effect is achieved on the different modeling tasks as far as possible, model parameters are prevented from being excessively fitted to a certain biased passenger group distribution, the situations that the model is difficult to maintain, modeling intermediate information is recycled or the model cannot be built and the like caused by respective modeling in the existing mode are avoided, and the technical problem that the realization difficulty of the existing credit risk modeling mode is large is solved.
Further, a second embodiment of the characteristic variable learning method of the present invention is proposed based on the above-described first embodiment shown in fig. 2. In the present embodiment, step S20 includes:
building a model comprising an input variable layer, a shared hidden layer, a shared output layer and a task output layer by using a neural network model architecture and a multi-task learning algorithm to serve as the target model;
and inputting the modeling characteristic variable set into an input variable layer of the target model, so as to transmit the modeling characteristic variable set from the input variable layer into the shared hidden layer, and performing deep interaction between variables in the shared hidden layer to obtain the parameter sharing result.
In this embodiment, as shown in fig. 3. The target model architecture comprises an input variable layer, a shared hidden layer, a shared output layer and a task output layer. It should be noted that the two-layer shared hidden layer in the figure does not represent the number of actual shared hidden layers. The input variable layer is used for inputting all modeling characteristic variables in the modeling characteristic variable set as the input of the input variable layer in the target model by the variable information input terminal, then the modeling characteristic variable set is transmitted from the input variable layer to the shared hidden layer, deep interaction among variables is carried out in the shared hidden layer, model parameters learn by relying on the gradient transmitted back from the upper layer, and the parameter sharing result is obtained.
Further, the step of transferring the modeling feature variable set from the input variable layer to the shared hidden layer, and performing deep interaction between variables in the shared hidden layer to obtain the parameter sharing result includes:
transmitting the modeling characteristic variable set from the input variable layer to the shared hidden layer so that the modeling characteristic variable set learns multiple groups of universal variable interaction information in the shared hidden layer;
and transmitting the multiple groups of the general variable interaction information into the shared output layer so as to summarize the multiple groups of the general variable interaction information into the parameter sharing result on the shared output layer.
In this embodiment, after all modeling feature variables in the modeling feature variable set are transmitted from the input variable layer to the shared hidden layer, modeling feature variables corresponding to different modeling tasks learn multiple groups of general variable interaction information in the shared hidden layer, and then are transmitted to the shared output layer for summarization, and the summarized result is used as the parameter sharing result.
Further, step S30 includes:
transmitting the parameter sharing result into the task output layer, and learning to obtain a gradient optimization direction based on a plurality of modeling tasks on the task output layer;
and optimizing the parameter sharing result by utilizing the gradient optimization direction to obtain the objective function based on the optimized parameter sharing result.
In this embodiment, as shown in fig. 3, the task output layer accesses a plurality of different modeling tasks (e.g., task object 1 to task object n in fig. 3) on the basis of sharing the parameter sharing result of the output layer, so as to further learn about the modeling tasks, and at the same time, the loss function in this layer transfers the gradient value between the output result and the modeling tasks downward, where the loss function is obtained by weighted summation of the target loss functions of the modeling tasks, and the weight is generally set by manual experience. The loss function is used to measure the degree of difference between the model f (x) and the modeling task y, with the derivative as the gradient direction for parametric optimization. In the learning stage in the task output layer, different target output results and training targets are input into the loss function to obtain a gradient optimization direction, and parameters of the lower layer shared part are optimized to obtain a final target function.
Further, in the embodiment, a parameter-sharing neural network model is constructed for different customer groups and business targets in different business stages, parameters are shared, output is shared, and parameter learning is influenced by accessing a plurality of targets, so that a better effect is achieved on each target as much as possible, and model parameters are prevented from being over-fitted to a certain biased customer group distribution.
Further, a third embodiment of the characteristic variable learning method of the present invention is proposed based on the first embodiment shown in fig. 2 described above. In the present embodiment, step S10 includes:
acquiring customer information and business stage information corresponding to the customer information, and dividing the customer information into a plurality of different customer groups according to the business stage information;
and sampling the modeling sample clients from a plurality of different client groups, and acquiring a modeling characteristic variable set corresponding to the modeling sample clients.
In this embodiment, the service stage information refers to identification information of different stages where the customer is in the credit service, such as account opening, first loan, stock, overdue, and the like, and different stages may be divided into different customer groups. The terminal classifies the existing whole client source according to different service stages as classification bases, and divides a plurality of client groups. And selecting part of customers from the multiple customer groups through sample sampling to serve as modeling sample customers, and finally correspondingly acquiring modeling characteristic variables of the modeling sample customers on a specified platform to serve as the modeling characteristic variable set. The specific sampling method may be random sampling or other sampling methods, and this embodiment is not limited in particular.
Further, the step of obtaining the customer information and the service phase information corresponding to the customer information, and dividing the customer information into a plurality of different customer groups according to the service phase information includes:
obtaining the client information and the service phase information, and carrying out desensitization treatment on the client information to obtain desensitization client information;
and carrying out guest group division on the desensitization guest information by using the service phase information to obtain a plurality of different guest groups.
In this embodiment, data desensitization refers to performing data deformation on some sensitive information according to a desensitization rule, so as to reliably protect sensitive private data. Under the condition of relating to client security data or some business sensitive data, the real data is modified and provided for test use under the condition of not violating system rules, and data desensitization is required to be carried out on personal information such as identification numbers, mobile phone numbers, card numbers, client numbers and the like. Since the client information may include sensitive data, after the terminal acquires the client information, it needs to perform desensitization processing on the client information to obtain desensitized client information, and then perform client group division.
Further, after step S30, the method further includes:
and determining a modeling task to be predicted in the modeling tasks, predicting and outputting a target output result of the modeling task to be predicted based on the target function.
In this embodiment, in a prediction stage after learning is completed, a user may select one or more target output nodes according to a requirement, and after receiving a currently determined modeling task to be predicted (i.e., a business target), a terminal may output a prediction result for the user to apply, so that the scheme is more convenient and flexible.
As shown in fig. 4, the present invention also provides a characteristic variable learning device including:
the characteristic variable acquisition module 10 is configured to acquire modeling characteristic variable sets corresponding to modeling sample clients in different business stages, where the modeling sample clients are obtained based on different client groups, and the modeling characteristic variable sets correspond to different modeling tasks;
the model parameter sharing module 20 is configured to build a target model based on a neural network model and a multi-task learning algorithm framework, and cooperate the modeling characteristic variable set as an input of the target model, so as to perform parameter sharing on the modeling characteristic variable set in the target model to obtain a parameter sharing result;
and the objective function learning module 30 is configured to perform parallel learning on the parameter sharing result for a plurality of modeling tasks to obtain an objective function through learning, where the objective function is used to uniformly represent a mapping relationship between the modeling characteristic variable set and the plurality of modeling tasks.
The method executed by each program module can refer to each embodiment of the characteristic variable learning method of the present invention, and is not described herein again.
The invention also provides a characteristic variable learning device.
The characteristic variable learning device comprises a processor, a memory and a characteristic variable learning program stored on the memory and capable of running on the processor, wherein the characteristic variable learning program realizes the steps of the characteristic variable learning method when being executed by the processor.
The method implemented when the characteristic variable learning program is executed may refer to each embodiment of the characteristic variable learning method of the present invention, and details thereof are not repeated herein.
The invention also provides a computer readable storage medium.
The computer-readable storage medium of the present invention has stored thereon a characteristic variable learning program which, when executed by a processor, implements the steps of the characteristic variable learning method as described above.
The method implemented when the characteristic variable learning program is executed may refer to each embodiment of the characteristic variable learning method of the present invention, and details thereof are not repeated herein.
It should be noted that, in this document, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or system 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 system. 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 system that comprises the element.
The above-mentioned serial numbers of the embodiments of the present invention are merely for description and do not represent the merits of the embodiments.
Through the above description of the embodiments, those skilled in the art will clearly understand that the method of the above embodiments can be implemented by software plus a necessary general hardware platform, and certainly can also be implemented by hardware, but in many cases, the former is a better implementation manner. Based on such understanding, the technical solution of the present invention may be embodied in the form of a software product, which is stored in a storage medium (e.g., ROM/RAM, magnetic disk, optical disk) as described above and includes instructions for enabling a terminal device (e.g., a mobile phone, a computer, a server, an air conditioner, or a network device) to execute the method according to the embodiments of the present invention.
The above description is only a preferred embodiment of the present invention, and not intended to limit the scope of the present invention, and all modifications of equivalent structures and equivalent processes, which are made by using the contents of the present specification and the accompanying drawings, or directly or indirectly applied to other related technical fields, are included in the scope of the present invention.
Claims (10)
1. A feature variable learning method, characterized by comprising:
the method comprises the steps of obtaining modeling characteristic variable sets corresponding to modeling sample clients in different business stages, wherein the modeling sample clients are obtained based on different client groups, and the modeling characteristic variable sets correspond to different modeling tasks;
building a target model based on a neural network model and a multi-task learning algorithm framework, and combining the modeling characteristic variable set as the input of the target model so as to perform parameter sharing on the modeling characteristic variable set in the target model to obtain a parameter sharing result;
and performing parallel learning on the parameter sharing result aiming at the modeling tasks to obtain an objective function through learning, wherein the objective function is used for uniformly representing the mapping relation between the modeling characteristic variable set and the modeling tasks.
2. The feature variable learning method according to claim 1, wherein the step of building a target model based on a neural network model and a multi-task learning algorithm architecture, and using the set of modeling feature variables as inputs of the target model to perform parameter sharing on the set of modeling feature variables in the target model to obtain a parameter sharing result comprises:
building a model comprising an input variable layer, a shared hidden layer, a shared output layer and a task output layer by using a neural network model architecture and a multi-task learning algorithm to serve as the target model;
and inputting the modeling characteristic variable set into an input variable layer of the target model, so as to transmit the modeling characteristic variable set from the input variable layer into the shared hidden layer, and performing deep interaction between variables in the shared hidden layer to obtain the parameter sharing result.
3. The feature variable learning method according to claim 2, wherein the step of passing the set of modeled feature variables from the input variable layer into the shared hidden layer and performing deep interaction between variables in the shared hidden layer to obtain the parameter sharing result comprises:
transmitting the modeling characteristic variable set from the input variable layer to the shared hidden layer so that the modeling characteristic variable set learns multiple groups of universal variable interaction information in the shared hidden layer;
and transmitting the multiple groups of the general variable interaction information into the shared output layer so as to summarize the multiple groups of the general variable interaction information into the parameter sharing result on the shared output layer.
4. The feature variable learning method according to claim 2, wherein the step of learning the parameter sharing result in parallel for a plurality of the modeling tasks to obtain an objective function comprises:
transmitting the parameter sharing result into the task output layer, and learning to obtain a gradient optimization direction based on a plurality of modeling tasks on the task output layer;
and optimizing the parameter sharing result by utilizing the gradient optimization direction to obtain the objective function based on the optimized parameter sharing result.
5. The feature variable learning method of claim 1, wherein the step of obtaining modeling feature variable sets corresponding to modeling sample clients in a plurality of different business phases comprises:
acquiring customer information and business stage information corresponding to the customer information, and dividing the customer information into a plurality of different customer groups according to the business stage information;
and sampling the modeling sample clients from a plurality of different client groups, and acquiring a modeling characteristic variable set corresponding to the modeling sample clients.
6. The feature variable learning method according to claim 5, wherein the step of obtaining the customer information and the business phase information corresponding to the customer information, and dividing the customer information into a plurality of different customer groups according to the business phase information comprises:
obtaining the client information and the service phase information, and carrying out desensitization treatment on the client information to obtain desensitization client information;
and carrying out guest group division on the desensitization guest information by using the service phase information to obtain a plurality of different guest groups.
7. The feature variable learning method according to any one of claims 1 to 6, wherein after the step of learning the parameter sharing result in parallel for several modeling tasks to obtain an objective function, further comprising:
and determining a modeling task to be predicted in the modeling tasks, predicting and outputting a target output result of the modeling task to be predicted based on the target function.
8. A characteristic variable learning device characterized by comprising:
the characteristic variable acquisition module is used for acquiring modeling characteristic variable sets corresponding to modeling sample clients in different business stages, wherein the modeling sample clients are obtained based on different client groups, and the modeling characteristic variable sets correspond to different modeling tasks;
the model parameter sharing module is used for building a target model based on a neural network model and a multi-task learning algorithm framework, and the modeling characteristic variable set is used as the input of the target model so as to perform parameter sharing on the modeling characteristic variable set in the target model to obtain a parameter sharing result;
and the target function learning module is used for performing parallel learning on the parameter sharing result aiming at the modeling tasks to obtain a target function through learning, wherein the target function is used for uniformly representing the mapping relation between the modeling characteristic variable set and the modeling tasks.
9. A characteristic variable learning device characterized by comprising: memory, processor and a feature variable learning program stored on the memory and executable on the processor, which when executed by the processor implements the steps of the feature variable learning method according to any one of claims 1 to 7.
10. A computer-readable storage medium, characterized in that the computer-readable storage medium has stored thereon a feature variable learning program which, when executed by a processor, implements the steps of the feature variable learning method according to any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110045805.8A CN112766490B (en) | 2021-01-13 | 2021-01-13 | Feature variable learning method, device, equipment and computer readable storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110045805.8A CN112766490B (en) | 2021-01-13 | 2021-01-13 | Feature variable learning method, device, equipment and computer readable storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112766490A true CN112766490A (en) | 2021-05-07 |
| CN112766490B CN112766490B (en) | 2024-05-24 |
Family
ID=75700461
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110045805.8A Active CN112766490B (en) | 2021-01-13 | 2021-01-13 | Feature variable learning method, device, equipment and computer readable storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112766490B (en) |
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110414624A (en) * | 2019-08-06 | 2019-11-05 | 广东工业大学 | Disaggregated model construction method and device based on multi-task learning |
| CN110443364A (en) * | 2019-06-21 | 2019-11-12 | 深圳大学 | A kind of deep neural network multitask hyperparameter optimization method and device |
| CN110598842A (en) * | 2019-07-17 | 2019-12-20 | 深圳大学 | Deep neural network hyper-parameter optimization method, electronic device and storage medium |
| CN110796166A (en) * | 2019-09-25 | 2020-02-14 | 浙江大学 | Attention mechanism-based multitask image processing method |
| US20200074341A1 (en) * | 2018-08-30 | 2020-03-05 | NEC Laboratories Europe GmbH | Method and system for scalable multi-task learning with convex clustering |
| CN111325357A (en) * | 2020-03-03 | 2020-06-23 | 深圳前海微众银行股份有限公司 | Machine learning modeling method, device, equipment and computer readable storage medium |
| US20200334520A1 (en) * | 2019-04-19 | 2020-10-22 | Microsoft Technology Licensing, Llc | Multi-task machine learning architectures and training procedures |
| CN112001484A (en) * | 2020-08-22 | 2020-11-27 | 哈尔滨工业大学 | Safety defect report prediction method based on multitask deep learning |
| CN112163676A (en) * | 2020-10-13 | 2021-01-01 | 北京百度网讯科技有限公司 | Multitask service prediction model training method, device, equipment and storage medium |
-
2021
- 2021-01-13 CN CN202110045805.8A patent/CN112766490B/en active Active
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20200074341A1 (en) * | 2018-08-30 | 2020-03-05 | NEC Laboratories Europe GmbH | Method and system for scalable multi-task learning with convex clustering |
| US20200334520A1 (en) * | 2019-04-19 | 2020-10-22 | Microsoft Technology Licensing, Llc | Multi-task machine learning architectures and training procedures |
| CN110443364A (en) * | 2019-06-21 | 2019-11-12 | 深圳大学 | A kind of deep neural network multitask hyperparameter optimization method and device |
| CN110598842A (en) * | 2019-07-17 | 2019-12-20 | 深圳大学 | Deep neural network hyper-parameter optimization method, electronic device and storage medium |
| CN110414624A (en) * | 2019-08-06 | 2019-11-05 | 广东工业大学 | Disaggregated model construction method and device based on multi-task learning |
| CN110796166A (en) * | 2019-09-25 | 2020-02-14 | 浙江大学 | Attention mechanism-based multitask image processing method |
| CN111325357A (en) * | 2020-03-03 | 2020-06-23 | 深圳前海微众银行股份有限公司 | Machine learning modeling method, device, equipment and computer readable storage medium |
| CN112001484A (en) * | 2020-08-22 | 2020-11-27 | 哈尔滨工业大学 | Safety defect report prediction method based on multitask deep learning |
| CN112163676A (en) * | 2020-10-13 | 2021-01-01 | 北京百度网讯科技有限公司 | Multitask service prediction model training method, device, equipment and storage medium |
Non-Patent Citations (3)
| Title |
|---|
| ZHENDONG ZHAO 等: "Geometry Embedded Multi-Task Network for Lane Detection and Classification", 《JOURNAL OF PHYSICS: CONFERENCE SERIES》, 23 August 2020 (2020-08-23), pages 1 - 13 * |
| 亢良伊;王建飞;刘杰;叶丹;: "可扩展机器学习的并行与分布式优化算法综述", 软件学报, no. 01, 9 October 2017 (2017-10-09), pages 113 - 134 * |
| 徐小峰;刘靖;: "多阶段并发的协同物流网络任务-资源集成调配模型", 运筹与管理, no. 07, 25 July 2018 (2018-07-25), pages 47 - 52 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112766490B (en) | 2024-05-24 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111310814A (en) | Method and device for training business prediction model by utilizing unbalanced positive and negative samples | |
| US9720940B2 (en) | Data analysis computer system and method for parallelized and modularized analysis of big data | |
| CN107169573A (en) | Using composite machine learning model come the method and system of perform prediction | |
| CN107169574A (en) | Using nested machine learning model come the method and system of perform prediction | |
| CN110070115A (en) | A kind of single pixel attack sample generating method, device, equipment and storage medium | |
| CN112785005B (en) | Multi-objective task assistant decision-making method and device, computer equipment and medium | |
| CN112328646B (en) | Multitask course recommendation method and device, computer equipment and storage medium | |
| CN116257363B (en) | Resource scheduling method, device, equipment and storage medium | |
| CN110969441A (en) | Anti-fraud model processing method and device based on deep learning | |
| CN111814117A (en) | Model interpretation method, device and readable storage medium | |
| Terziyan et al. | Causality-aware convolutional neural networks for advanced image classification and generation | |
| Matinkhah et al. | Emerging artificial intelligence application: reinforcement learning issues on current internet of things | |
| CN112669084A (en) | Policy determination method, device and computer-readable storage medium | |
| Carneiro et al. | Using genetic algorithms to create solutions for conflict resolution | |
| CN112766490B (en) | Feature variable learning method, device, equipment and computer readable storage medium | |
| CN114627085A (en) | Target image identification method and device, storage medium and electronic equipment | |
| Sharma | ENFORCEMENT EXPLORATION OF FUZZY LOGISTICS ACUITIES IN DIGITAL MARKETING DECISION AMBITIOUS TRIALS | |
| Alidoust | AGI brain II: the upgraded version with increased versatility index | |
| Sarang et al. | Estimators | |
| Lazzeri | Adopting Machine Learning and AI for Data-driven Decisions | |
| Sheng et al. | VPnet: A Vulnerability Prioritization Approach Using Pointer Network and Deep Reinforcement Learning | |
| Lazzeri | Understanding Algorithms and the Business Questions that Algorithms can answer Defining Business Metrics and Business Impact | |
| CN114297840A (en) | Modeling method, system, terminal device and storage medium of sensitivity scoring model | |
| CN114169541A (en) | Model construction method, device, equipment and computer readable storage medium | |
| Khurshid et al. | Performance Analysis of Students in Educational Big Data using Deep Learning |
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 | ||
| GR01 | Patent grant |