CN115907955A

CN115907955A - Grading card model construction method, device, equipment, storage medium and product

Info

Publication number: CN115907955A
Application number: CN202211148694.4A
Authority: CN
Inventors: 伏峰
Original assignee: China Construction Bank Corp; CCB Finetech Co Ltd
Current assignee: China Construction Bank Corp; CCB Finetech Co Ltd
Priority date: 2022-09-21
Filing date: 2022-09-21
Publication date: 2023-04-04

Abstract

The application discloses a scoring card model construction method, a scoring card model construction device, scoring card model construction equipment, a scoring card model construction storage medium and a scoring card model construction product. The method for constructing the scoring card model comprises the following steps: acquiring an input variable set of the first scoring card model and input variable sets corresponding to the plurality of second scoring card models respectively; combining and de-duplicating variables in each input variable set to generate an input variable total set; vectorizing each element in the input variable total set according to the position of the model input variable in the variable total set to generate an n-dimensional vector of the model; respectively calculating the similarity between every two different models based on the n-dimensional vectors corresponding to the models to obtain a similarity matrix; respectively calculating the weight of each model based on the similarity matrix; and performing weighted calculation on the first scoring card model and the plurality of second scoring card models based on the weight of each model to obtain a target scoring card model. According to the embodiment of the application, the problems of high uncertainty and low accuracy of the grading card model result can be solved.

Description

Grading card model construction method, device, equipment, storage medium and product

Technical Field

The application belongs to the field of data processing, and particularly relates to a scoring card model construction method, device, equipment, storage medium and product.

Background

Currently, the scoring card modeling technology is widely used in the financial industry. For example, the user's credit is scored using a credit scoring model.

At present, a construction mode of a rating card model is that a model is constructed by using the experience of an expert, the model is subjective and depends on the long-term service knowledge accumulation of the expert, and the model has high uncertainty and low accuracy.

Disclosure of Invention

The embodiment of the application provides a method, a device, equipment, a storage medium and a product for constructing a rating card model, so as to improve the certainty and accuracy of the rating card model constructed according to expert experience.

In a first aspect, an embodiment of the present application provides a score card model construction method, including:

acquiring an input variable set of a first scoring card model and input variable sets corresponding to a plurality of second scoring card models respectively, wherein the first scoring card model is a scoring card model constructed according to expert experience, and the second scoring card model is a scoring card model obtained by performing model training according to training sample data;

merging and de-duplicating the input variables in each input variable set to generate an input variable total set;

the following operations are performed separately for each model: vectorizing each element in the input variable total set according to the position of the input variable of the model in the input variable total set to generate an n-dimensional vector of the model, wherein n is a positive integer;

respectively calculating the similarity between every two different models based on the n-dimensional vectors corresponding to the models to obtain a similarity matrix, wherein the rows and the columns of the similarity matrix are the identifications corresponding to the models;

respectively calculating the weight of each model based on the similarity matrix;

and performing weighted calculation on the first scoring card model and the plurality of second scoring card models based on the weight of each model to obtain a target scoring card model.

In a second aspect, an embodiment of the present application provides a score card model building apparatus, including:

the acquisition module is used for acquiring an input variable set of the first scoring card model and input variable sets corresponding to the plurality of second scoring card models respectively;

the processing module is used for merging and de-duplicating the input variables in each input variable set to generate an input variable total set;

the vector generation module is used for vectorizing each element in the input variable total set according to the position of the input variable of the model in the variable total set to generate an n-dimensional vector of the model;

the similarity calculation module is used for calculating the similarity between every two different models respectively based on the n-dimensional vectors corresponding to the models to obtain a similarity matrix;

the weight calculation module is used for calculating the weight of each model respectively based on the similarity matrix;

and the weighting calculation module is used for carrying out weighting calculation on the first scoring card model and the plurality of second scoring card models based on the weight of each model to obtain a target scoring card model.

In a third aspect, an embodiment of the present application provides an electronic device, including: a processor and a memory storing computer program instructions;

the processor, when executing the computer program instructions, performs the steps of the scorecard model construction method as in any of the embodiments of the first aspect.

In a fourth aspect, the present application provides a computer storage medium, where computer program instructions are stored on the computer readable storage medium, and when executed by a processor, the computer program instructions implement the steps of the score card model building method in any one of the embodiments of the first aspect.

In a fifth aspect, the present application provides a computer program product, and when instructions in the computer program product are executed by a processor of an electronic device, the electronic device executes the steps of the score card model building method in any one of the embodiments of the first aspect.

According to the scoring card model construction method, device, equipment, storage medium and product, the target scoring card model is obtained by correcting the first scoring card model through the plurality of second scoring card models. The first scoring card model is a scoring card model constructed according to expert experience, and the second scoring card model is a scoring card model obtained by performing model training according to training sample data. The second scoring card model is obtained by performing model training according to training sample data, so that the accuracy and the certainty of the second scoring card model are high. Therefore, the accuracy and the certainty of the target scoring card model obtained by correcting the first scoring card model by using the plurality of second scoring card models are higher, and the problems of higher uncertainty and lower accuracy of the existing scoring card model constructed according to expert experience are solved.

Drawings

In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings needed to be used in the embodiments of the present application will be briefly described below, and for those skilled in the art, other drawings can be obtained according to the drawings without creative efforts.

Fig. 1 is a schematic flow chart of a scoring card model construction method according to an embodiment of the present application;

FIG. 2 is a flowchart illustrating a specific implementation manner of step S130;

FIG. 3 is a flowchart illustrating an embodiment of step S140;

FIG. 4 is a flowchart illustrating a specific implementation manner of step S150;

fig. 5 is a schematic flow chart of a scoring card model construction method according to another embodiment of the present application;

fig. 6 is a schematic structural diagram of a scoring card model building device according to an embodiment of the present application;

fig. 7 is a schematic structural diagram of an electronic device according to an embodiment of the present application.

Detailed Description

Features and exemplary embodiments of various aspects of the present application will be described in detail below, and in order to make objects, technical solutions and advantages of the present application more apparent, the present application will be further described in detail below with reference to the accompanying drawings and specific embodiments. It should be understood that the specific embodiments described herein are intended to be illustrative only and are not intended to be limiting. It will be apparent to one skilled in the art that the present application may be practiced without some of these specific details. The following description of the embodiments is merely intended to provide a better understanding of the present application by illustrating examples thereof.

It is noted that, herein, relational terms such as first and second, and the like may be used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Also, 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 … …" does not exclude the presence of another like element in a process, method, article, or apparatus that comprises the element.

According to the technical scheme, the data acquisition, storage, use, processing and the like meet the relevant regulations of national laws and regulations.

At present, a construction mode of a rating card model is that the model is constructed by using the experience of an expert, and the model is subjective and depends on the long-term service knowledge accumulation of the expert, so that the model has high uncertainty and low accuracy.

In order to solve the problems of the prior art, the embodiment of the application provides a scoring card model construction method, device, equipment, storage medium and product. First, a method for constructing a score card model provided in the embodiment of the present application is described below.

Fig. 1 is a flowchart illustrating a scoring card model construction method according to an embodiment of the present application. As shown in fig. 1, the method for constructing a score card model may specifically include the following steps:

s110, acquiring an input variable set of a first scoring card model and input variable sets corresponding to a plurality of second scoring card models respectively, wherein the first scoring card model is a scoring card model constructed according to expert experience, and the second scoring card model is a scoring card model obtained by performing model training according to training sample data;

s120, merging and de-duplicating the input variables in each input variable set to generate an input variable total set;

s130, respectively executing the following operations for each model: vectorizing each element in the input variable total set according to the position of the input variable of the model in the variable total set to generate an n-dimensional vector of the model, wherein n is a positive integer;

s140, respectively calculating the similarity between every two different models based on the n-dimensional vectors corresponding to the models to obtain a similarity matrix, wherein the rows and the columns of the similarity matrix are all the models;

s150, respectively calculating the weight of each model based on the similarity matrix;

and S160, carrying out weighted calculation on the first scoring card model and the plurality of second scoring card models based on the weight of each model to obtain a target scoring card model.

Thus, the target scoring card model is obtained by correcting the first scoring card model by using the plurality of second scoring card models. The first scoring card model is a scoring card model constructed according to expert experience, and the second scoring card model is a scoring card model obtained by performing model training according to training sample data. The second scoring card model is obtained by performing model training according to training sample data, so that the accuracy and the certainty of the second scoring card model are high. Therefore, the accuracy and the certainty of the target scoring card model obtained by correcting the first scoring card model by using the plurality of second scoring card models are higher, and the problems of higher uncertainty and lower accuracy of the existing scoring card model constructed according to expert experience are solved.

Specific implementations of the above steps are described below.

In some embodiments, in S110, variables in the input variable sets of the first scoring card model and the plurality of second scoring card models according to the embodiments of the present application are evaluation indexes used for constructing the scoring card model.

As one example, the variables in the input variable sets of the first scoring card model may include age, occupation, age, place of work, and industry to which work belongs, and the variables in the input variable sets of the second scoring card model may include age, occupation, presence of room, presence of car, presence of marriage, gender, scholarly, height, weight, age of work, place of work, and industry to which work belongs.

In some embodiments, in S120, a union of the input variable sets may be calculated, and then the elements in the union may be subjected to deduplication processing, so as to obtain a total input variable set.

As an example, for example, the input variable set a of the first rating card model is { age, occupation, working age, working location, and working industry }, the input variable set B of the plurality of second rating card models is { age, occupation, presence or absence of a room, presence or absence of a car, presence or absence of a wedding }, the set C is { sex, calendar, presence or absence of a room, presence or absence of a car, presence or absence of a wedding }, the set D is { height, weight, working age, working location, and working industry }, and then the sets are combined and deduplicated, thereby obtaining the input variable total set E of { age, occupation, working age, working location, working industry, presence or absence of a room, presence or absence of a wedding, sex, calendar, height, and weight }.

In order to facilitate the calculation of the subsequent similarity, in S130, the following operations are respectively performed for each model: and vectorizing each element in the input variable total set according to the position of the input variable of the model in the variable total set to generate an n-dimensional vector of the model, wherein n is a positive integer.

It should be noted that a value of n is determined according to the number of elements in the variable total set, and the value of n is not less than the number of elements in the variable total set.

As an example, in order to further improve the calculation efficiency, as shown in fig. 2, the S130 may specifically include:

s131, setting an element at the position of an input variable of the model in the input variable total set as a first numerical value;

s132, setting elements of the model in the input variable total set, which are not located at the positions, as second numerical values;

s133, according to the position of the input variable of the model in the input variable total set, vectorizing each element in the input variable total set to generate an n-dimensional vector of the model.

It should be noted that, the first value and the second value have different values, and in order to simplify the calculation process, the first value may be 1, and the second value may be 0.

When the input variable sets of the first model are respectively { age, occupation, working age, working location, industry of work }, { age, occupation, whether there is a room, whether there is a car, whether there is a wedding }, { sex, academic calendar, whether there is a room, whether there is a car, whether there is a wedding } and { height, weight, working age, working location, and industry of work }, the generated input variable sets are { age, occupation, working age, working location, industry of work, whether there is a room, whether there is a car, whether there is a wedding, sex, academic calendar, height, and weight }, the input variable sets of the first model are respectively corresponding to {11,1,1,1,0,0,0,0,0,0,0}, {1,1,0,0,0,1,1,1,0,0,0,0}, {0,0,0,0,0,1,1,1,1,1,0,0} and {0,0,1,1,1,0,0,0,0,0,1,1}.

In some embodiments, in S140, based on the n-dimensional vector corresponding to each model, the similarity between each two different models is respectively calculated, so as to obtain a model similarity matrix table as shown in table 1, where rows and columns of the similarity matrix are identifiers of each model. Wherein, A ₁ 、A ₂ 、A ₃ 、A _i 、A _j 、A _m For model identification, a _(i,j) To representModel A _i And model A _j And two characteristics are satisfied: a is _(i,j) ＝a _(j,i) I.e. model A _i And model A _j Similarity of (2) to model A _j And model A _i The similarity is the same; a is a _(i,i) =1, i.e. the similarity value of the model to itself is 1.

TABLE 1 model similarity matrix table

Because the cosine distance can well represent the similarity between two objects, the cosine distance is adopted to represent the similarity of two models in the application. As an example, as shown in fig. 3, the S140 may specifically include:

s141, respectively calculating cosine distances between n-dimensional vectors of every two different models;

illustratively, the cosine distance between the n-dimensional vectors of each two different models can be calculated separately using the following formula (1):

wherein A is _i And B _i Is an n-dimensional vector of two different models, sim is the cosine distance.

And S142, performing matrixing on each cosine distance to obtain a similarity matrix.

Based on the cosine distances calculated by the formula (1), the cosine distances are placed at the corresponding positions of the similarity matrix to realize matrixing, and a similarity matrix table of the specific embodiment shown in table 2 is obtained.

TABLE 2 exemplary similarity matrix Table

	T	M1	M2	M3
					T	1	0.4	0	0.6
M1	0.4	1	0.6	0
					M2	0	0.6	1	0
M3	0.6	0	0	1

Where T is a specific example of the first rating card model, and M1, M2, and M3 are specific examples of the second rating card model, respectively.

In some embodiments, in S150, the weight of each model is calculated separately based on the similarity matrix.

In order to improve the calculation efficiency, as shown in fig. 4, the step S150 may specifically include:

s151, summarizing the similarity between every two different models to obtain the sum of the similarities;

s152, summing the similarity of the first model and other models to obtain the similarity of the first model, wherein the first model is any one of the first scoring card model and the plurality of second scoring card models, and the other models are models except the first model in any one of the first scoring card model and the plurality of second scoring card models;

s153, calculating the ratio of the similarity of the first model to the total similarity of the first model to obtain the weight of the first model.

As an example, the similarity of the first model and the other models is summed up using the following formula (2):

wherein, a _i,j Representing the similarity of the respective models, S _i The sum of the model similarities is shown, and m represents the number of similarities.

Calculating the ratio of the similarity of the first model to the sum of the similarities by adopting the following formula (3) to obtain the weight of the first model:

W _i ＝S _i /(S ₁ +S ₂ +…+S _i +…+S _m ) (3)

wherein, W _i Weight, S, representing the first model ₁ 、S ₂ 、S _i 、S _m Representing the similarity of the respective models.

In some embodiments, in S160, the first scoring card model and the plurality of second scoring card models are weighted and calculated based on the weight of each model, so as to obtain the target scoring card model.

As an example, the product of the similarity of the target scoring card model and the corresponding weight is calculated, the products of the similarity of the plurality of second scoring card models and the corresponding weight are calculated, and the obtained products are added and summed to obtain the final target scoring card model. Specifically, the following formula (4) is adopted for weighting calculation:

R＝W ₁ *A ₁ +W ₂ *A ₂ +…+W _i *A _i +…+W _n *A _n (4)

wherein R represents a value after the weighted calculation, A ₁ Represents the corresponding score value of the target score card model, A ₂ 、A _i 、A _j Represents the corresponding score value, W, of the second scorecard model ₁ 、W ₂ 、W ₃ 、W _i 、W _n Representing the original weight of each model.

In order to verify the accuracy of the obtained model, as another embodiment of the scoring card model construction method according to the present application, as shown in fig. 5, after the step S160, the method may further include:

s510, obtaining a test data set of test input variables, wherein the test input variables comprise input variables in an input variable total set, and the test data set comprises multiple groups of data of the input variables;

s520, inputting the multiple groups of data into a first scoring card model, a second scoring card model and a target scoring card model respectively to obtain a first scoring set, a second scoring set and a third scoring set respectively;

s530, according to the first scoring set, the second scoring set, the third scoring set, the first scoring card model and the second scoring card model, KS inspection is carried out on the target scoring card model respectively to obtain the maximum vertical difference between scoring accumulated distribution curves in the first scoring set, the second scoring set, the third scoring set, the first scoring card model and the second scoring card model;

s540, carrying out normalization processing on each maximum vertical difference to obtain the maximum vertical difference after normalization processing;

and S550, adjusting the target scoring card model according to the maximum vertical difference after normalization processing to obtain a final target scoring card model.

As an example, KS tests were performed on the target score card models, respectively, as shown in table 3, resulting in the maximum vertical difference between the respective score cumulative distribution curves:

TABLE 3KS inspection Table

A ₁

A ₂

A ₃

…

A _j

…

A _m

R

D ₁

D ₂

D ₃

…

D _j

…

D _m

Wherein D is ₁ 、D ₂ 、D ₃ 、D _j 、D _m Representing the respective maximum vertical differences, R represents the target scoring card model, A ₁ 、A ₂ 、A ₃ 、A _j 、A _m Representing the respective scoring sets, the first scoring card model and the second scoring card model.

Normalizing each maximum vertical difference by using the following formula (5) to obtain a normalized maximum vertical difference:

E _i ＝D _i /(D ₁ +D ₂ +D ₃ +…+D _i +…+D _m ) (5)

wherein D is ₁ 、D ₂ 、D ₃ 、D _i 、D _m Denotes the respective maximum vertical difference, E _i The value after the normalization process for the maximum vertical difference is shown.

And (3) according to the maximum vertical difference after the normalization processing, adjusting the target score card model by adopting the following formula (6):

wherein, F _i Indicates the weight adjustment factor, has a default value of 0, and can be set to [0,1 ]]And (4) adjusting within the range, wherein the sum of all involved weight adjusting coefficients is 1.

Therefore, after the target scoring card model is obtained, the test data set of the input variable set is obtained, and the target scoring card model is adjusted through a weighting method because the test data set comprises multiple groups of data of the input variables, so that the accuracy of the model is verified.

It should be noted that the application scenario described in the embodiment of the present application is for more clearly explaining the technical solution of the embodiment of the present application, and does not constitute a limitation on the technical solution provided in the embodiment of the present application. As can be known to those skilled in the art, with the emergence of new application scenarios, the technical solutions provided in the embodiments of the present application are also applicable to similar technical problems.

Based on the same inventive concept, the application also provides a score card model construction device, which is specifically described in detail with reference to fig. 6.

Fig. 6 is a schematic structural diagram of a score card model building apparatus according to an embodiment of the present application.

As shown in fig. 6, the score card model building apparatus 600 may include:

an obtaining module 610, configured to obtain an input variable set of the first scoring card model and input variable sets corresponding to the plurality of second scoring card models, respectively;

a processing module 620, configured to merge and perform deduplication processing on input variables in each input variable set, to generate an input variable total set;

the vector generating module 630 is configured to vectorize, according to a position of an input variable of the model in a variable total set, each element in the input variable total set, and generate an n-dimensional vector of the model;

a similarity calculation module 640, configured to calculate, based on the n-dimensional vectors corresponding to the models, similarities between each two different models respectively, so as to obtain a similarity matrix;

a weight calculation module 650, configured to calculate a weight of each model based on the similarity matrix;

and the weighting calculation module 660 is configured to perform weighting calculation on the first scoring card model and the plurality of second scoring card models based on the weight of each model to obtain a target scoring card model.

The scoring card model building apparatus 600 is described in detail below, and is specifically as follows:

in some embodiments, in order to facilitate data acquisition, the acquisition module 610 may further include:

the first obtaining unit is used for obtaining an input variable set of the first scoring card model and input variable sets corresponding to the plurality of second scoring card models respectively;

and the second acquisition unit is used for acquiring a test data set of the test input variable.

In some embodiments, in order to facilitate processing the acquired data, the processing module 620 may further include the following units:

the first processing unit is used for merging the input variables in each input variable set;

and the second processing unit is used for carrying out duplicate removal on the combined total set to generate an input variable total set.

In some embodiments, in order to vectorize each element in the total set of input variables, the vector generation module 630 may further include the following units:

the first number setting unit is used for setting an element at the position of an input variable of the model in the input variable total set as a first numerical value;

and the second number setting unit is used for setting the elements of the input variables of the model in the input variable total set, which are not at the positions, as second numerical values.

In some embodiments, to facilitate calculating the similarity, the similarity calculating module 640 may further include the following units:

the cosine distance calculation unit is used for calculating the cosine distance between the n-dimensional vectors of every two different models respectively;

and the matrixing unit is used for matrixing each cosine distance to obtain a similarity matrix.

In some embodiments, to facilitate calculating the weight, the weight calculating module 650 may further include the following units:

the summarizing unit is used for summarizing the similarity between every two different models to obtain the sum of the similarity;

the adding unit is used for adding the similarity of the first model and other models to obtain the similarity of the first model;

and the weight calculation unit is used for calculating the ratio of the similarity of the first model to the total similarity to obtain the weight of the first model.

In some embodiments, in order to obtain the target score card model and verify the accuracy of the model, the weighting calculation module 660 may further include the following units:

the scoring set acquisition unit is used for respectively inputting the test data sets into the first scoring card model, the second scoring card model and the target scoring card model to respectively obtain a first scoring set, a second scoring set and a third scoring set;

the KS testing unit is used for respectively testing the target scoring card model according to the first scoring set, the second scoring set, the third scoring set, the first scoring card model and the second scoring card model to obtain the maximum vertical difference between the scoring accumulated distribution curves in the first scoring set, the second scoring set, the third scoring set, the first scoring card model and the second scoring card model;

and the adjusting unit is used for carrying out normalization processing on each maximum vertical difference to obtain the maximum vertical difference after the normalization processing, and adjusting the target scoring card model according to the maximum vertical difference after the normalization processing to obtain the final target scoring card model.

The electronic device 700 may include a processor 720 and memory 730 that stores computer program instructions.

Specifically, the processor 720 may include a Central Processing Unit (CPU), or an Application Specific Integrated Circuit (ASIC), or may be configured to implement one or more Integrated circuits of the embodiments of the present Application.

Memory 730 may include mass storage for data or instructions. By way of example, and not limitation, memory 730 may include a Hard Disk Drive (HDD), a floppy Disk Drive, flash memory, an optical Disk, a magneto-optical Disk, tape, or a Universal Serial Bus (USB) Drive or a combination of two or more of these. Storage 730 may include removable or non-removable (or fixed) media, where appropriate. Memory 730 may be internal or external to the integrated gateway disaster recovery device, where appropriate. In a particular embodiment, the memory 330 is non-volatile solid-state memory.

In a particular embodiment, memory 730 includes Read Only Memory (ROM). Where appropriate, the ROM may be mask-programmed ROM, programmable ROM (PROM), erasable PROM (EPROM), electrically Erasable PROM (EEPROM), electrically rewritable ROM (EAROM), or flash memory, or a combination of two or more of these. The memory may include Read Only Memory (ROM), random Access Memory (RAM), magnetic disk storage media devices, optical storage media devices, flash memory devices, electrical, optical, or other physical/tangible memory storage devices. Thus, in general, the memory includes one or more tangible (non-transitory) computer-readable storage media (e.g., memory devices) encoded with software comprising computer-executable instructions and when the software is executed (e.g., by one or more processors), it is operable to perform operations described with reference to the methods according to an aspect of the present disclosure.

Processor 720 may implement any of the above described scoring card model construction methods by reading and executing computer program instructions stored in memory 730.

In one example, electronic device 700 may also include a communication interface 740 and a bus 710. As shown in fig. 7, the processor 720, the memory 730, and the communication interface 740 are connected via a bus 710 to perform communication with each other.

The communication interface 740 is mainly used for implementing communication between modules, apparatuses, units and/or devices in the embodiments of the present application.

Bus 710 includes hardware, software, or both to couple the components of the scorecard model building apparatus to each other. By way of example, and not limitation, a bus may include an Accelerated Graphics Port (AGP) or other graphics bus, an Enhanced Industry Standard Architecture (EISA) bus, a Front Side Bus (FSB), a Hypertransport (HT) interconnect, an Industry Standard Architecture (ISA) bus, an infiniband interconnect, a Low Pin Count (LPC) bus, a memory bus, a Micro Channel Architecture (MCA) bus, a Peripheral Component Interconnect (PCI) bus, a PCI-Express (PCI-X) bus, a Serial Advanced Technology Attachment (SATA) bus, a video electronics standards association local (VLB) bus, or other suitable bus or a combination of two or more of these. Bus 710 may include one or more buses, where appropriate. Although specific buses are described and shown in the embodiments of the application, any suitable buses or interconnects are contemplated by the application.

Illustratively, the electronic device 700 may be a mobile phone, a tablet computer, a notebook computer, a palm top computer, a vehicle-mounted electronic device, an ultra-mobile personal computer (UMPC), a netbook or a Personal Digital Assistant (PDA), and the like.

The electronic device 700 may execute the scoring card model construction method in the embodiment of the present application, so as to implement the scoring card model construction method and apparatus described in conjunction with fig. 1 and fig. 6.

In addition, in combination with the scoring card model construction method in the foregoing embodiment, the embodiment of the present application may provide a computer storage medium to implement. The computer storage medium having computer program instructions stored thereon; the computer program instructions, when executed by a processor, implement any one of the score card model construction methods in the above embodiments. Examples of computer readable storage media include non-transitory computer readable storage media such as portable disks, hard disks, random Access Memories (RAMs), read Only Memories (ROMs), erasable programmable read only memories (EPROMs or flash memories), portable compact disc read only memories (CD-ROMs), optical storage devices, magnetic storage devices, and so forth.

It is to be understood that the present application is not limited to the particular arrangements and instrumentalities described above and shown in the attached drawings. A detailed description of known methods is omitted herein for the sake of brevity. In the above embodiments, several specific steps are described and shown as examples. However, the method processes of the present application are not limited to the specific steps described and illustrated, and those skilled in the art can make various changes, modifications, and additions or change the order between the steps after comprehending the spirit of the present application.

The functional blocks shown in the above structural block diagrams may be implemented as hardware, software, firmware, or a combination thereof. When implemented in hardware, it may be, for example, an electronic circuit, an Application Specific Integrated Circuit (ASIC), suitable firmware, plug-in, function card, or the like. When implemented in software, the elements of the present application are the programs or code segments used to perform the required tasks. The program or code segments may be stored in a machine-readable medium or transmitted by a data signal carried in a carrier wave over a transmission medium or a communication link. A "machine-readable medium" may include any medium that can store or transfer information. Examples of a machine-readable medium include an electronic circuit, a semiconductor memory device, a ROM, a flash memory, an Erasable ROM (EROM), a floppy disk, a CD-ROM, an optical disk, a hard disk, an optical fiber medium, a Radio Frequency (RF) link, and so forth. The code segments may be downloaded via computer networks such as the internet, intranets, etc.

It should also be noted that the exemplary embodiments mentioned in this application describe some methods or systems based on a series of steps or devices. However, the present application is not limited to the order of the above-described steps, that is, the steps may be performed in the order mentioned in the embodiments, may be performed in an order different from the order in the embodiments, or may be performed simultaneously.

Aspects of the present application are described above with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems) and computer program products according to embodiments of the application. It will be understood that each block of the flowchart illustrations and/or block diagrams, and combinations of blocks in the flowchart illustrations 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, 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, enable the implementation of the functions/acts specified in the flowchart and/or block diagram block or blocks. Such a processor may be, but is not limited to, a general purpose processor, a special purpose processor, an application specific processor, or a field programmable logic circuit. It will also be understood that each block of the block diagrams and/or flowchart illustration, and combinations of blocks in the block diagrams and/or flowchart illustration, can be implemented by special purpose hardware for performing the specified functions or acts, or combinations of special purpose hardware and computer instructions.

As will be apparent to those skilled in the art, for convenience and brevity of description, the specific working processes of the systems, modules and units described above may refer to corresponding processes in the foregoing method embodiments, and are not described herein again. It should be understood that the scope of the present application is not limited thereto, and any person skilled in the art can easily conceive various equivalent modifications or substitutions within the technical scope of the present application, and these modifications or substitutions should be covered within the scope of the present application.

Claims

1. A method for constructing a scoring card model is characterized by comprising the following steps:

the following operations are performed separately for each model: vectorizing each element in the input variable total set according to the position of an input variable of a model in the input variable total set to generate an n-dimensional vector of the model, wherein n is a positive integer;

respectively calculating the similarity between every two different models based on the n-dimensional vectors corresponding to the models to obtain a similarity matrix, wherein the rows and the columns of the similarity matrix are respectively the identifications corresponding to the models;

2. The method of claim 1, wherein vectorizing each element in the total set of input variables of the model according to their position in the total set of input variables, to generate an n-dimensional vector of the model, comprises:

setting the element of the position of the input variable of the model in the input variable total set as a first numerical value, and setting the element of the position of the input variable of the model in the input variable total set as a second numerical value to obtain an n-dimensional vector of the model.

3. The method according to claim 2, wherein the calculating the similarity between each two different models based on the n-dimensional vectors corresponding to the models respectively to obtain a similarity matrix comprises:

respectively calculating cosine distances between n-dimensional vectors of every two different models;

and performing matrixing on each cosine distance to obtain a similarity matrix.

4. The method of claim 1, wherein separately calculating the weight of each model based on the similarity matrix comprises:

summarizing the similarity between every two different models to obtain the sum of the similarities;

the following operations are respectively executed for the first model: the similarity of the first model and other models is added to obtain the similarity of the first model;

calculating the ratio of the similarity of the first model to the sum of the similarities to obtain the weight of the first model;

the first model is any one of the first scoring card model and the plurality of second scoring card models, and the other models are models other than the first model in any one of the first scoring card model and the plurality of second scoring card models.

5. The method of claim 1, wherein after said obtaining a target scoring card model, the method further comprises:

acquiring a test data set of the test input variables, wherein the test input variables comprise input variables in the variable total set; the test dataset comprises a plurality of groups of data of input variables;

and respectively inputting the multiple groups of data into a first scoring card model, a second scoring card model and a target scoring card model to respectively obtain a first scoring set, a second scoring set and a third scoring set.

6. The method of claim 5, wherein after said obtaining the first, second, and third sets of scores, respectively, the method further comprises:

performing KS inspection on the target scoring card model according to the first scoring set, the second scoring set, the third scoring set, the first scoring card model and the second scoring card model to obtain the maximum vertical difference between scoring accumulated distribution curves in the first scoring set, the second scoring set, the third scoring set, the first scoring card model and the second scoring card model;

normalizing each maximum vertical difference to obtain a normalized maximum vertical difference;

and adjusting the target scoring card model according to the maximum vertical difference after the normalization processing to obtain a final target scoring card model.

7. The method according to claim 6, wherein the adjusting the target score card model according to the maximum vertical difference after the normalization processing to obtain a final target score card model comprises:

adjusting the target scoring card model according to the following formula to obtain a final target scoring card model:

wherein, E _i Is the maximum vertical difference after the normalization process, F _i Represents a weight adjustment coefficient, W _i Representing the original weight of each scoring card model, A _i And expressing the corresponding score value of each scoring card model.

8. An apparatus for constructing a score card model, comprising:

9. An electronic device, characterized in that the electronic device comprises: a processor and a memory storing computer program instructions;

the processor, when executing the computer program instructions, implements the steps of the scorecard model construction method according to any one of claims 1 to 7.

10. A computer-readable storage medium, having computer program instructions stored thereon, which, when executed by a processor, implement the steps of the scorecard model construction method according to any one of claims 1 to 7.

11. A computer program product, characterized in that instructions in the computer program product, when executed by a processor of an electronic device, cause the electronic device to perform the steps of the scoring card model construction method according to any one of claims 1 to 7.