Customer classification method and system

Abstract

The invention provides a customer classification method and system, which can be used in the financial field or other fields, and the method comprises the steps: obtaining mechanism feature data of a financial mechanism and feature data of a target feature type corresponding to a to-be-processed customer; determining a classification result of the to-be-processed customer according to mechanism feature data of the financial mechanism, feature data of the target feature type and a preset classification model; the target feature type is obtained by applying a first feature screening mode or a second feature screening mode in advance; the preset classification model is obtained by pre-training a fuzzy clustering algorithm based on the feature data set. According to the invention, the efficiency and reliability of customer classification can be improved, and the reliability of subsequent deposit business processing can be improved.

Description

Technical Field

[0001] The present application relates to the field of artificial intelligence technology, and in particular to a customer classification method and system. Background Art

[0002] At present, the bank's unit time deposit products are mainly divided according to the deposit term, mainly including three months, six months, one year, two years, three years and five years. The interest of each product is fixed according to the bank's listed interest rate.

[0003] However, with the continuous deepening of interest rate marketization, in order to adapt to market competition, maintain and expand high-quality customers, and meet the interest rate needs of different customers, the account managers of some banks manually classify deposit customers, determine customer qualifications, and decide whether to provide interest rate discounts based on customer qualifications, as well as the specific floating rate or points to complete deposit business processing.

[0004] Customer classification requires manual intervention, which is time-consuming, inefficient, and involves operational risks. Summary of the invention

[0005] In response to at least one problem in the prior art, the present application proposes a customer classification method and system, which can improve the efficiency and reliability of customer classification, and further improve the reliability of subsequent deposit business processing.

[0006] In order to solve the above technical problems, this application provides the following technical solutions:

[0007] In a first aspect, the present application provides a customer classification method, comprising:

[0008] Obtaining the institutional characteristic data of the financial institution and the characteristic data of the target characteristic type corresponding to the customer to be processed;

[0009] Determining the classification result of the customer to be processed according to the institutional characteristic data of the financial institution, the characteristic data of the target characteristic type and a preset classification model;

[0010] The target feature type is obtained by pre-applying the first feature screening method or the second feature screening method; the preset classification model is obtained by pre-training a fuzzy clustering algorithm based on a feature data set;

[0011] The first feature screening method includes: applying the original feature data groups corresponding to the plurality of feature types to determine the correlation coefficients corresponding to the various feature types; performing feature screening on the various feature types based on the correlation coefficients to obtain the first intermediate feature type; applying the original feature data groups corresponding to the first intermediate feature type to determine the feature contribution of the first intermediate feature type; performing feature screening on the first intermediate feature type based on the feature contribution to obtain the target feature type;

[0012] The second feature screening method includes: applying the original feature data groups corresponding to each feature type to determine the correlation coefficient and feature contribution of each feature type; performing feature screening on each feature type based on the correlation coefficient to obtain a first intermediate feature type; performing feature screening on each feature type based on the feature contribution to obtain a second intermediate feature type; and determining a feature type that belongs to both the first intermediate feature type and the second intermediate feature type as the target feature type.

[0013] In one embodiment, before acquiring the institutional characteristic data of the financial institution and the characteristic data of the target characteristic type corresponding to the customer to be processed, the method further includes:

[0014] Acquire an original feature data set, the original feature data set comprising: a plurality of original samples, each original sample comprising: original feature data corresponding to a plurality of feature types; divide the original feature data set into a plurality of groups of original feature data groups, the original feature data in the same group of original feature data groups correspond to the same feature type, and the feature types corresponding to the original feature data in different groups of original feature data groups are different;

[0015] Based on the original feature data groups corresponding to each of the feature types, the first feature screening method or the second feature screening method is applied to screen out the target feature type from each of the feature types.

[0016] In one embodiment, the application of the original feature data groups corresponding to each feature type to determine the correlation coefficient and feature contribution of each feature type includes:

[0017] Obtain the actual classification results corresponding to each original sample;

[0018] The classification algorithm is trained using each original sample and its corresponding actual classification results to obtain a preset initial classification model;

[0019] Input each original feature data in the same original feature data group into a preset initial classification model to obtain a prediction classification result corresponding to the original feature data;

[0020] Determine the actual classification result corresponding to the original sample to which the original feature data belongs as the actual classification result corresponding to the original feature data;

[0021] According to the actual classification results and predicted classification results corresponding to each original feature data in the same original feature data group, the feature contribution of the feature type corresponding to the original feature data group is determined; and the correlation coefficient of each feature type is determined by applying the original feature data groups corresponding to each feature type.

[0022] In one embodiment, before determining the classification result of the customer to be processed according to the institution characteristic data of the financial institution, the characteristic data of the target characteristic type and the preset classification model, the method further includes:

[0023] Acquire a feature data set, the feature data set comprising: a plurality of samples, each sample comprising: historical institution feature data and historical feature data of a target feature type;

[0024] The feature data set is used to train a fuzzy clustering algorithm to obtain the classification model.

[0025] In one embodiment, before acquiring the institutional characteristic data of the financial institution and the characteristic data of the target characteristic type corresponding to the customer to be processed, the method further includes:

[0026] receiving a deposit request from the pending customer;

[0027] Correspondingly, after determining the classification result of the customer to be processed according to the institutional characteristic data of the financial institution, the characteristic data of the target characteristic type and the preset classification model, the method further includes:

[0028] If the classification result is a high-quality customer, then determine the current interest rate increase ratio corresponding to the customer to be processed according to the deposit request;

[0029] Determine the final execution interest rate of the customer to be processed based on the pre-obtained listed interest rate and the current interest rate increase ratio;

[0030] The deposit business processing process corresponding to the deposit request is completed according to the final execution interest rate.

[0031] In one embodiment, determining the current interest rate increase ratio corresponding to the pending customer according to the deposit request includes:

[0032] Determine the total deposit interval and the type of region corresponding to the customer to be processed according to the pre-acquired average daily deposit base corresponding to the customer to be processed, the deposit amount and the location information in the deposit request;

[0033] Determine the interest rate increase ratio corresponding to the customer to be processed according to the preset interest rate increase ratio model, the pre-acquired average daily deposit base, the classification result, the total deposit range and the type of region;

[0034] Among them, the preset interest rate increase ratio model is obtained by pre-training the regression model using batch samples and their corresponding real interest rate increase ratios, and each sample includes: the historical customer's average daily deposit base, classification results, deposit range and region type.

[0035] In one embodiment, determining the current interest rate increase ratio corresponding to the pending customer according to the deposit request includes:

[0036] Determine the deposit ratio of the customer to be processed according to the pre-acquired total account balance, the daily average deposit base and the deposit amount in the deposit request corresponding to the customer to be processed;

[0037] Determine the deposit corresponding interval and interest rate increase ratio interval corresponding to the customer to be processed according to the preset corresponding relationship table, the pre-acquired daily average deposit base and the deposit ratio, wherein the preset corresponding relationship table includes: the corresponding relationship between multiple daily average deposit base intervals, deposit corresponding intervals and interest rate increase intervals;

[0038] The current interest rate increase ratio corresponding to the customer to be processed is determined according to the deposit ratio, the deposit corresponding range and the interest rate increase ratio range.

[0039] In a second aspect, the present application provides a customer classification system, comprising:

[0040] An acquisition module, used to acquire the institutional characteristic data of the financial institution and the characteristic data of the target characteristic type corresponding to the customer to be processed;

[0041] A determination module, used to determine the classification result of the customer to be processed according to the institutional characteristic data of the financial institution, the characteristic data of the target characteristic type and a preset classification model;

[0042] The target feature type is obtained by pre-applying the first feature screening method or the second feature screening method; the preset classification model is obtained by pre-training a fuzzy clustering algorithm based on a feature data set;

[0043] The first feature screening method includes: applying the original feature data groups corresponding to the plurality of feature types to determine the correlation coefficients corresponding to the various feature types; performing feature screening on the various feature types based on the correlation coefficients to obtain the first intermediate feature type; applying the original feature data groups corresponding to the first intermediate feature type to determine the feature contribution of the first intermediate feature type; performing feature screening on the first intermediate feature type based on the feature contribution to obtain the target feature type;

[0044] The second feature screening method includes: applying the original feature data groups corresponding to each feature type to determine the correlation coefficient and feature contribution of each feature type; performing feature screening on each feature type based on the correlation coefficient to obtain a first intermediate feature type; performing feature screening on each feature type based on the feature contribution to obtain a second intermediate feature type; and determining a feature type that belongs to both the first intermediate feature type and the second intermediate feature type as the target feature type.

[0045] In one embodiment, the customer classification system further includes:

[0046] The data set acquisition module is used to acquire an original feature data set, wherein the original feature data set includes: a plurality of original samples, each of which includes: original feature data corresponding to a plurality of feature types; the original feature data set is divided into a plurality of groups of original feature data groups, wherein the feature types corresponding to the original feature data in the same group of original feature data groups are the same, and the feature types corresponding to the original feature data in different groups of original feature data groups are different;

[0047] The screening module is used to apply the first feature screening method or the second feature screening method based on the original feature data groups corresponding to each feature type, to screen out the target feature type from each feature type.

[0048] In one embodiment, the application of the original feature data groups corresponding to each feature type to determine the correlation coefficient and feature contribution of each feature type includes:

[0049] Obtain the actual classification results corresponding to each original sample;

[0050] The classification algorithm is trained using each original sample and its corresponding actual classification results to obtain a preset initial classification model;

[0051] Input each original feature data in the same original feature data group into a preset initial classification model to obtain a prediction classification result corresponding to the original feature data;

[0052] Determine the actual classification result corresponding to the original sample to which the original feature data belongs as the actual classification result corresponding to the original feature data;

[0053] According to the actual classification results and predicted classification results corresponding to each original feature data in the same original feature data group, the feature contribution of the feature type corresponding to the original feature data group is determined; and the correlation coefficient of each feature type is determined by applying the original feature data groups corresponding to each feature type.

[0054] In one embodiment, the customer classification system further includes:

[0055] A sample acquisition module is used to acquire a feature data set, the feature data set comprising: a plurality of samples, each sample comprising: historical institution feature data and historical feature data of a target feature type;

[0056] The training module is used to train the fuzzy clustering algorithm using the feature data set to obtain the classification model.

[0057] In one embodiment, the customer classification system further includes:

[0058] A receiving module, used for receiving the deposit request of the client to be processed;

[0059] Correspondingly, the customer classification system further includes:

[0060] A module for determining an interest rate increase ratio is used to determine the current interest rate increase ratio corresponding to the customer to be processed according to the deposit request if the classification result is a high-quality customer;

[0061] A module for determining a final execution ratio is used to determine the final execution ratio of the customer to be processed according to the pre-acquired listing rate and the current interest rate increase ratio;

[0062] A processing module is used to complete the deposit business processing process corresponding to the deposit request according to the final execution interest rate.

[0063] In one embodiment, the module for determining the interest rate increase ratio includes:

[0064] A region type determination unit, configured to determine a total deposit interval and a region type corresponding to the customer to be processed according to the pre-acquired average daily deposit base corresponding to the customer to be processed, the deposit amount and the location information in the deposit request;

[0065] A determination ratio unit is used to determine the interest rate increase ratio corresponding to the customer to be processed based on a preset interest rate increase ratio model, the pre-acquired average daily deposit base, the classification result, the total deposit range and the type of region; wherein the preset interest rate increase ratio model is obtained by pre-training the regression model using batch samples and their respective corresponding real interest rate increase ratios, and each sample includes: the average daily deposit base of historical customers, the classification result, the deposit range and the type of region.

[0066] In one embodiment, the module for determining the interest rate increase ratio includes:

[0067] The first determining unit is used to determine the deposit ratio of the customer to be processed according to the pre-acquired total account balance corresponding to the customer to be processed, the average daily deposit base and the deposit amount in the deposit request.

[0068] The second determination unit is used to determine the deposit corresponding interval and interest rate floating ratio interval corresponding to the customer to be processed according to a preset corresponding relationship table, the pre-acquired average daily deposit base and the deposit ratio, and the preset corresponding relationship table includes: the correspondence between multiple average daily deposit base intervals, deposit corresponding intervals and interest rate floating intervals.

[0069] The third determining unit is used to determine the current interest rate increase ratio corresponding to the customer to be processed according to the deposit ratio, the deposit corresponding interval and the interest rate increase ratio interval.

[0070] In a third aspect, the present application provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the customer classification method when executing the program.

[0071] In a fourth aspect, the present application provides a computer-readable storage medium having computer instructions stored thereon, which implement the customer classification method when the instructions are executed.

[0072] It can be seen from the above technical solution that the present application provides a customer classification method and system. The method includes:

[0073] Obtain the institutional characteristic data of the financial institution and the characteristic data of the target characteristic type corresponding to the customer to be processed; determine the classification result of the customer to be processed according to the institutional characteristic data of the financial institution, the characteristic data of the target characteristic type and the preset classification model; the target characteristic type is obtained by pre-applying the first characteristic screening method or the second characteristic screening method; the preset classification model is obtained by pre-training the fuzzy clustering algorithm based on the characteristic data set; the application of the first characteristic screening method includes: applying the original characteristic data groups corresponding to each of the multiple characteristic types to determine the correlation coefficient corresponding to each characteristic type; performing characteristic screening on each characteristic type based on the correlation coefficient to obtain a first intermediate characteristic type; applying the original characteristic data group corresponding to the first intermediate characteristic type to determine the characteristic contribution of the first intermediate characteristic type; performing characteristic screening on the first intermediate characteristic type based on the characteristic contribution to obtain to the target feature type; the second feature screening method includes: applying the original feature data group corresponding to each feature type to determine the correlation coefficient and feature contribution of each feature type; based on the correlation coefficient, feature screening is performed on each feature type to obtain the first intermediate feature type; based on the feature contribution, feature screening is performed on each feature type to obtain the second intermediate feature type; the feature type belonging to both the first intermediate feature type and the second intermediate feature type is determined as the target feature type, which can improve the efficiency and reliability of customer classification, and thus improve the reliability of subsequent deposit business processing; specifically, the customer classification can be performed on the public customers, and the interest rate floating ratio can be effectively obtained according to the customer's average daily deposit number and deposit growth for the daily deposit scenario of high-quality customers, so as to improve the deposit efficiency of the customer's interest rate floating; while ensuring the customer's deposit income, it can effectively encourage the customer to deposit more and get more. For banks, accurately and reasonably floating interest rates for customers can improve the ability to acquire and retain customers, and at the same time can effectively reduce human operations and interventions, improve processing efficiency, and reduce operational risks. In addition, a maximum floating ratio is also set to effectively control the bank's deposit costs. In addition to being applicable to corporate time deposits, it can also be applied to other deposit products such as demand deposits, notice deposits, and large-denomination certificates of deposit, comprehensively improving customers' deposit experience and bank competitiveness. This makes it more reliable, flexible, and efficient for large corporate customers to increase interest rates when depositing different time products. In combination with the application scenarios of interest rate adjustments, institutional feature data is introduced into customer classification to improve the reliability of the impact of customer classification results on subsequent interest rate adjustments; while providing customers with deposit returns, it also improves customer stickiness and competitiveness for banks and ensures capital sedimentation. BRIEF DESCRIPTION OF THE DRAWINGS

[0074] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0075] Figure 1 It is a logical diagram of a time deposit processing process in an example of the prior art;

[0076] Figure 2 is a flow chart of a customer classification method in an embodiment of the present application;

[0077] Figure 3 is a flow chart of a customer classification method in another embodiment of the present application;

[0078] Figure 4 It is a logical diagram of a customer classification method in an example of the present application;

[0079] Figure 5 is a schematic diagram of the structure of a customer classification system in an embodiment of the present application;

[0080] Figure 6 A schematic block diagram of the system structure of an electronic device according to an embodiment of the present application. DETAILED DESCRIPTION

[0081] In order to enable those skilled in the art to better understand the technical solutions in this specification, the technical solutions in the embodiments of this application will be clearly and completely described below in conjunction with the drawings in the embodiments of this application. Obviously, the described embodiments are only part of the embodiments of this application, not all of them. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0082] However, with the continuous deepening of interest rate marketization, in order to adapt to market competition, maintain and expand high-quality customers, and meet the interest rate needs of different customers, some banks have signed interest rate increase agreements for corporate customers through offline negotiations with corporate customers, stipulating the points or proportion of interest rate increase to improve deposit-taking capacity.

[0083] At present, the steps for corporate customers to purchase corporate time deposit products are mainly divided into the following steps:

[0084] Step 1: The customer negotiates with the bank and applies to sign an interest rate increase agreement.

[0085] Step 2: The bank account manager decides whether to provide an interest rate discount and the specific increase ratio or points based on the customer's qualifications.

[0086] Step 3: Customers purchase regular products through various channels.

[0087] Step 4: The system determines whether the customer has an interest rate increase agreement. If there is no agreement, the interest will be calculated based on the listed interest rate of the product. If there is an agreement, the interest will be calculated based on the listed interest rate plus the interest rate discount.

[0088] Through the above steps, it can be seen that the method of determining the fixed floating point or ratio of the customer interest rate through offline negotiation with corporate customers lacks flexibility and scientificity, has a long process, and requires time and cost, and is inefficient. In addition, since the bank account manager needs to decide whether to provide interest rate discounts and the specific floating ratio or points based on experience and customer qualifications, there is also operational risk. The process of a fixed deposit in an example of the prior art is as follows: Figure 1 Based on this, the embodiment of the present application proposes a customer classification method and system, which can improve the efficiency and reliability of customer classification, and thus improve the reliability of subsequent deposit business processing, and can solve the problem that the interest rate adjustment method is not flexible enough and different interest rates cannot be provided for each customer, and automatically increase the interest rate for corporate customers, meeting the needs of providing different interest rate discounts for high-quality corporate customers when selecting various deposit products.

[0089] It should be noted that the customer classification method and system disclosed in this application can be used in the field of financial technology, and can also be used in any field other than financial technology. The application field of the customer classification method and system disclosed in this application is not limited. The acquisition, storage, use, and processing of data in the technical solutions of each embodiment of this application comply with the relevant provisions of laws and regulations.

[0090] The details are described through the following embodiments.

[0091] In order to improve the efficiency and reliability of customer classification, and thus improve the reliability of subsequent deposit business processing, this embodiment provides a customer classification method in which the execution subject is a customer classification system, and the customer classification system includes but is not limited to a server, such as Figure 2 As shown, the method specifically includes the following contents:

[0092] Step 100: Obtain the institutional characteristic data of the financial institution and the characteristic data of the target characteristic type corresponding to the customer to be processed.

[0093] Step 200: Determine the classification result of the customer to be processed according to the institutional characteristic data of the financial institution, the characteristic data of the target characteristic type and the preset classification model; the target characteristic type is obtained by pre-applying the first characteristic screening method or the second characteristic screening method; the preset classification model is obtained by pre-training the fuzzy clustering algorithm based on the characteristic data set; the application of the first characteristic screening method includes: applying the original characteristic data groups corresponding to each of the plurality of characteristic types to determine the correlation coefficient corresponding to each characteristic type; performing characteristic screening on each characteristic type based on the correlation coefficient to obtain a first intermediate characteristic type; applying the original characteristic data groups corresponding to the first intermediate characteristic type to determine the characteristic contribution of the first intermediate characteristic type; performing characteristic screening on the first intermediate characteristic type based on the characteristic contribution to obtain the target characteristic type; the application of the second characteristic screening method includes: applying the original characteristic data groups corresponding to each characteristic type to determine the correlation coefficient and characteristic contribution of each characteristic type; performing characteristic screening on each characteristic type based on the correlation coefficient to obtain the first intermediate characteristic type; performing characteristic screening on each characteristic type based on the characteristic contribution to obtain the second intermediate characteristic type; and determining the characteristic type belonging to both the first intermediate characteristic type and the second intermediate characteristic type as the target characteristic type.

[0094] Specifically, the customer to be processed may be a corporate customer of the bank; the classification result of the customer to be processed may be one of a high-quality customer, a potential customer, a general customer, and a risky customer. The institutional characteristic data may include: the current total deposit amount of the financial institution; the institutional characteristic data may also include: the current deposit tendency index level of the financial institution, which may be divided into multiple deposit tendency index levels in advance according to actual needs, from expecting customers to deposit to not recommending customers to deposit.

[0095] Specifically, in order to improve the efficiency of customer classification, the original historical feature data corresponding to multiple categories of features can be obtained; feature selection can be performed on each category of features based on the correlation coefficient to obtain the first intermediate feature; feature screening can be performed based on the feature contribution of each category of the first intermediate feature to obtain the second intermediate feature, and the second intermediate feature can be determined as the target feature; feature data corresponding to the target feature can be obtained from the original historical feature data and determined as the historical feature data; the feature corresponding to the feature data of the customer to be processed can be the target feature. Furthermore, in order to further improve the efficiency of feature screening, while obtaining the first intermediate feature, feature screening can also be performed based on the feature contribution of each category of features to obtain the second intermediate feature, and the feature that belongs to both the first intermediate feature and the second intermediate feature can be determined as the target feature. The feature contribution of each category of features can be obtained based on the original historical feature data corresponding to multiple categories of features and the tree model.

[0096] Specifically, the customer to be processed may be a corporate customer; the target feature types corresponding to the feature data may include: asset size, number of employees, industry score, registration time, total assets, number of commonly used products, account transaction frequency, contribution, risk status, and risk status of customer-related companies. The asset size can be divided into the first to eighth levels; the total asset situation can include: the current market value of each asset; the number of commonly used financial products; contribution = total customer income - total customer cost = customer deposit benefit + customer loan benefit + customer intermediary business benefit, customer deposit benefit = customer deposit internal funds transfer income - customer deposit interest expenditure, customer deposit internal funds transfer income can be obtained by multiplying the average daily balance of each type of deposit of the customer by the funds transfer price, customer loan benefit = customer loan interest income - loan risk reserve - loan business tax and surcharge - internal funds transfer expenditure, customer intermediary business benefit = customer intermediary business fee income - customer intermediary business fee expenditure - business tax surcharge on intermediary business income; the asset size, number of employees, industry score, registration time, total assets, number of commonly used products, account transaction frequency, contribution, risk status and risk status of customer-related enterprises of the customers to be processed can all be pre-stored in the local database of the customer classification system.

[0097] In order to further improve the efficiency of screening and obtaining target feature types, and thus improve the efficiency of subsequent customer classification, before step 100, the following steps are further included:

[0098] Acquire an original feature data set, the original feature data set comprising: a plurality of original samples, each original sample comprising: original feature data corresponding to a plurality of feature types; divide the original feature data set into a plurality of original feature data groups, the original feature data in the same original feature data group corresponding to the same feature type, and the feature types corresponding to different original feature data groups being different; based on the original feature data groups corresponding to each feature type, apply the first feature screening method or the second feature screening method to screen out the target feature type from each feature type.

[0099] In order to further improve the accuracy and intelligence of determining the feature contribution, in one embodiment of the present application, the original feature data groups corresponding to each feature type are applied to determine the correlation coefficient and feature contribution of each feature type, including:

[0100] Obtain the actual classification results corresponding to each original sample; use each original sample and its corresponding actual classification results to train the classification algorithm to obtain a preset initial classification model; input each original feature data in the same original feature data group into the preset initial classification model to obtain the predicted classification result corresponding to the original feature data; determine the actual classification result corresponding to the original sample to which the original feature data belongs as the actual classification result corresponding to the original feature data; determine the feature contribution of the feature type corresponding to the original feature data group according to the actual classification result and the predicted classification result corresponding to each original feature data in the same original feature data group; and use the original feature data groups corresponding to each feature type to determine the correlation coefficient of each feature type.

[0101] For example, assuming that the original feature data in the original sample 1 includes: {x1, x2, x3}, and the corresponding classification result is K1; the original feature data in the original sample 2 includes: {y1, y2, y3}, and the corresponding classification result is K2; the original feature data group A is {x1, y1}, and the original feature data group B is {x2, y2}; {x1, 0}, {0, y1}, {x2, 0}, {0, y2} are respectively input into the preset initial classification model to obtain the respective predicted classification results R1, R2, R3, R4. The similarity mean (S1+S2) / 2 obtained according to the similarity S1 between R1 and K1 and the similarity S2 between R2 and K2 can be determined as the contribution of the feature type corresponding to the original feature data group A; the similarity mean (S3+S4) / 2 obtained according to the similarity S3 between R3 and K1 and the similarity S4 between R4 and K2 can be determined as the contribution of the feature type corresponding to the original feature data group B. The method for obtaining the feature contribution of the first intermediate feature type is similar.

[0102] In order to improve the reliability of classification model training, and then apply reliable classification models to improve the accuracy of customer classification, such as Figure 3 As shown, in one embodiment of the present application, before step 200, it also includes:

[0103] Step 021: Acquire a feature data set, which includes: multiple samples, each sample includes: historical institution feature data and historical feature data of a target feature type.

[0104] Specifically, the historical characteristic data may be characteristic data of historical customers.

[0105] Step 022: Use the feature data set to train a fuzzy clustering algorithm to obtain the classification model.

[0106] Specifically, the historical feature data can be standardized, and the feature values ​​of the standardized historical feature data can be screened through quantitative feature recognition, Pearson correlation coefficient, and feature contribution, and the main feature values ​​can be extracted. The fuzzy clustering algorithm is trained using the main feature values ​​to screen out the category with the highest score and evaluate it as a high-quality customer. Furthermore, the classification results of the customers to be processed and the information of the recommended financial products corresponding to the classification results can also be sent to the terminal device of the business personnel to improve the success rate of product recommendations.

[0107] In order to improve the flexibility of determining the final execution rate on the basis of improving the efficiency and reliability of customer classification, in one embodiment of the present application, before step 100, the following is further included:

[0108] Step 001: receiving the deposit request of the customer to be processed; correspondingly, after step 200, it also includes:

[0109] Step 300: If the classification result is a high-quality customer, then determine the current interest rate increase ratio corresponding to the customer to be processed based on the deposit request.

[0110] Specifically, a deposit request initiated by a client of a pending customer may be received; the deposit request may be a time deposit request; the deposit request may include: the deposit amount, deposit term and location information corresponding to the current deposit of the pending customer; the deposit term mainly includes three months, six months, one year, two years, three years and five years.

[0111] Step 400: Determine the final execution interest rate of the customer to be processed based on the pre-acquired listing interest rate and the current interest rate increase ratio.

[0112] Specifically, the final implementation interest rate = listed interest rate × (1 + current interest rate increase ratio).

[0113] Step 500: completing the deposit business processing process corresponding to the deposit request according to the final execution interest rate.

[0114] In order to improve the flexibility of determining the current interest rate increase ratio and improve the customer acquisition and retention capabilities, in one embodiment of the present application, the step 300 of determining the current interest rate increase ratio corresponding to the pending customer according to the classification result, the deposit request and the characteristic data includes:

[0115] Step 301: Determine the deposit ratio of the customer to be processed according to the pre-acquired total account balance, the daily average deposit base and the deposit amount in the deposit request corresponding to the customer to be processed.

[0116] Specifically, the average daily deposit base can be obtained by adding up the average daily deposits of all accounts opened by the customer last year; the deposit ratio P of the customer to be processed can be determined according to the following formula: d :

[0117]

[0118] Among them, d1 represents the deposit amount, r represents the total account balance, and x m Represents the average daily deposit base.

[0119] Step 302: Determine the deposit corresponding interval and interest rate increase ratio interval corresponding to the customer to be processed according to a preset correspondence table, the pre-acquired average daily deposit base and the deposit ratio, wherein the preset correspondence table includes: the correspondence between multiple average daily deposit base intervals, deposit corresponding intervals and interest rate increase ratio intervals.

[0120] Specifically, the deposit corresponding interval where the deposit ratio is located can be determined as the target deposit interval, and the daily average deposit base interval where the daily average deposit base is located can be determined as the target daily average deposit base interval; from the correspondence between the daily average deposit base interval, the deposit corresponding interval, and the interest rate increase interval, the interest rate increase interval corresponding to the target daily average deposit base interval and the target deposit interval can be determined as the interest rate increase ratio interval corresponding to the customer to be processed. The correspondence between the daily average deposit base interval, the deposit corresponding interval, and the interest rate increase interval can be shown in Table 1. For example, if x2<x m <x n , x 22 <p d <x 23 , where x m represents the average daily deposit base, p d It indicates the deposit ratio, and the interest rate increase interval of the customer to be processed is y2-y3 as shown in Table 1. The correspondence between each daily average deposit base interval, deposit corresponding interval and interest rate increase interval can be pre-stored in the customer classification system locally according to actual conditions, and this application does not limit this.

[0121] Table 1

[0122]

[0123] In one example, the correspondence between the daily average deposit base range, the deposit corresponding range and the interest rate floating range can be shown in Table 2.

[0124] Table 2

[0125]

[0126] Step 303: Determine the current interest rate increase ratio corresponding to the customer to be processed according to the deposit ratio, the deposit corresponding interval and the interest rate increase ratio interval.

[0127] Specifically, the current interest rate increase ratio corresponding to the customer to be processed can be determined according to the following formula:

[0128]

[0129] Among them, p d Indicates the deposit ratio of customers to be processed, x mn-1 represents the lower limit of the deposit interval of the pending customer, x mn Indicates the upper limit of the deposit interval of the pending customer, y n Indicates the upper limit of the interest rate increase ratio range for the customer to be processed, y n-1 Indicates the lower limit of the interest rate increase ratio range for the customers to be processed.

[0130] In order to improve the accuracy and automation of determining the interest rate increase ratio, in one embodiment of the present application, the step 300 of determining the interest rate increase ratio corresponding to the pending customer according to the deposit request includes: determining the total deposit range and the type of region corresponding to the pending customer according to the pre-acquired average daily deposit base corresponding to the pending customer, the deposit amount and the location information in the deposit request; determining the interest rate increase ratio corresponding to the pending customer according to a preset interest rate increase ratio model, the pre-acquired average daily deposit base, the classification result, the total deposit range and the type of region; wherein the preset interest rate increase ratio model is obtained by pre-training a regression model using batch samples and their respective corresponding real interest rate increase ratios, and each sample includes: the average daily deposit base of historical customers, the classification result, the deposit range and the type of region.

[0131] Specifically, multiple deposit intervals can be pre-set according to actual conditions, such as 0-10,000, 10,000-100,000, 100,000 and above, etc. The sum of the pre-acquired daily average deposit base corresponding to the pending customer and the deposit amount is determined as the target deposit amount, and the deposit interval where the target deposit amount is located is determined as the total deposit interval corresponding to the pending customer; the type of location corresponding to the location information can be determined from the correspondence between the preset location type and the location, and the correspondence between the preset location type and the location can be set according to actual conditions, and this application does not limit this. The location type can include: high-frequency trading areas and low-frequency trading areas. The preset interest rate floating ratio model can be as follows:

[0132] y=β0+β1×x+β2×α1+β2×α2×x+β3×α3

[0133] Wherein, y represents the interest rate increase ratio, x represents the total deposit range; α1 represents the classification result, α1=1 represents high-quality customers, α1=0 represents other types of customers; α2 represents the deposit term; α3 represents the type of region, α3=1 represents a high-frequency trading region, α3=0 represents a low-frequency trading region; β0 to β3 represent preset parameters, and batch samples and their corresponding labels can be used to fit the preset interest rate increase ratio model to obtain β0 to β3. Each sample includes: the average daily deposit base of historical customers, classification results, deposit range and type of region, and the label can be the actual interest rate increase ratio corresponding to the sample.

[0134] In order to improve the flexibility and automation of determining the current interest rate increase ratio when the deposit ratio exceeds the upper limit of the deposit corresponding interval, and effectively control the bank's deposit cost, in one embodiment of the present application, the step 300 of determining the current interest rate increase ratio corresponding to the customer to be processed according to the classification result, the deposit request and the characteristic data includes:

[0135] The deposit ratio of the customer to be processed is determined based on the pre-acquired total account balance, the average daily deposit base and the deposit amount in the deposit request corresponding to the customer to be processed; if the deposit ratio exceeds the upper limit of the corresponding interval of each deposit, the highest upper limit in each interest rate increase interval is determined as the current interest rate increase ratio.

[0136] Specifically, if The current interest rate increase ratio p i =λ.

[0137] In order to improve the flexibility and automation of determining the current interest rate increase ratio when the deposit ratio is less than zero, in one embodiment of the present application, step 300 determines the current interest rate increase ratio corresponding to the customer to be processed based on the classification result, the deposit request and the characteristic data, including: determining the deposit ratio of the customer to be processed based on the pre-acquired total account balance, the average daily deposit base and the deposit amount in the deposit request corresponding to the customer to be processed; if the deposit ratio is less than zero, determining the lower limit of the previous deposit corresponding interval of the highest deposit corresponding interval as the current interest rate increase ratio.

[0138] Specifically, if Then p i =y n-1 .

[0139] In order to further illustrate this solution, this application provides a specific application example of a customer classification method, which is described as follows:

[0140] First, using the fuzzy clustering method, the sample set is selected as U = {U1, ..., Un}, each sample has j features, denoted as Ui = (ui1, ..., uij); i = 1, 2, ..., n; uij represents the number of the jth feature describing the sample Ui. The j features include customer asset size, number of employees, customer industry characteristics, customer registration time, customer total assets, customer products, account transaction frequency, customer contribution, risk status, and customer-related enterprise risk status.

[0141] Among them, the customer asset size can be obtained from the ASSETAMOUNT asset size in the ECIS_M_TB1020 legal person customer information table, which is divided into 8 levels.

[0142] The number of employees can be obtained from the number of employees of NO_OF_EMP company in the ECIS_M_TB1020 legal person customer information table.

[0143] The characteristics of the industry in which the customer is located can be obtained by linking the INDUSTRYCODE industry code in the ECIS_M_TB1020 legal person customer information table with the TA602005 industry rating total score table in the CS2002 reporting system to obtain the customer's industry score.

[0144] The customer registration time can be obtained from the legal person start date ORG_START_DATE in the ECIS_M_TB1020 legal person customer information table.

[0145] The total customer assets are summarized according to the customer's cis number, and the assets APPRAISE_DATA1 in the ECIS_M_TB1311_1 customer assets and liabilities statistics table, the current market value ISS_VALUE of each asset in the TB2418 customer other asset information table, and the TB2417 customer intangible asset information table.

[0146] The products used by customers can be calculated from the ECIS_M_TB1311_1 Customer Assets and Liabilities Statistics Table, based on the number of products held in APPRAISE_FLAG2 and the years of use of the products in APPRAISE_FLAG1, to calculate the number of products commonly used by customers.

[0147] The account transaction frequency can be obtained from the public customer account relationship table ECIS_M_TB1560 and the account transaction details table NTHCHDT1.

[0148] Customer contribution can be calculated using the contribution calculation model, which is: customer contribution = total customer income - total customer cost = customer deposit efficiency + customer loan efficiency + customer intermediary business efficiency.

[0149] Among them, customer deposit benefits = customer deposit internal funds transfer income - customer deposit interest expenses. Customer deposit internal funds transfer income can be obtained by multiplying the average daily balance of each type of customer deposit by the funds transfer price. The data can be processed and obtained from the MID_D_NTHFDSUB customer deposit information table.

[0150] Customer loan benefit = customer loan interest income - loan risk reserve - loan business tax and surcharge - internal funds transfer expenditure. Customer loan interest income can be obtained from the LCB_NTHLIDTL corporate loan interest details table to obtain the interest income of various types of customer loan products. Other costs can be obtained from the T_KTBWKCLC_RGS loan fund cost statistics table.

[0151] Customer intermediary business benefits = customer intermediary business fee income - customer intermediary business fee expenditure - business tax surcharge on intermediary business income. The income and expenses generated by various customer products can be summarized from the intermediary business income table and expense table.

[0152] The customer risk status can be obtained from the credit risk assessment of the ECIS_M_TB1020 legal person customer information table CR_RSK_RATING. The risk status of the customer's associated enterprises can be obtained from the association query of the ECIS_M_TB1150 legal person customer and customer relationship table and the ECIS_M_TB1020 legal person customer information table.

[0153] According to the above method, the required data is collected from the bank's related reports and data warehouse system, and stored in submodule 1 to form a matrix X, where i = 1, ..., 5, j = 1, ..., n.

[0154]

[0155] The original data is standardized by using the standard deviation standardization method. Then, the preprocessed data set is screened by quantitative feature recognition, Pearson correlation coefficient, and feature contribution to extract the main feature values, and trained by machine learning algorithms. The comprehensive score of the samples is calculated, and finally each sample is systematically clustered to screen out the category with the highest score and rated as a high-quality customer. Other categories can be divided into potential customers, general customers, and risky customers according to the score, and high-quality customers are stored in submodule 2. The data in submodule 1 is automatically updated in batches regularly to prepare for changes in customer categories when important feature values ​​of customers change. When encountering a new customer, the customer's comprehensive score needs to be recalculated through the above model, and it is necessary to re-determine whether the customer is a high-quality customer. When a customer initiates a self-service deposit, if the customer is an existing customer and there is no significant change in the feature values ​​recently, it can be directly judged through submodule 2 whether the customer is a high-quality customer, and then determine whether to enter the self-service interest rate floating stage.

[0156] Scenario 1: An existing customer plans to make a 1-year fixed deposit of d1 yuan.

[0157] Step 11: Confirm whether the existing customer is a high-quality customer stored in submodule 2. If the customer is a high-quality customer, the interest rate can be directly increased for the customer using the interest rate increase preferential device. If the customer is another type of customer, the account manager needs to intervene to further review the customer's qualifications, and no automatic interest rate increase will be performed this time.

[0158] Step 12: Assuming that the existing customer is a high-quality customer in submodule 2, when the customer deposits money through the self-service channel, the interest rate floating device is enabled. Specifically, the stored historical interest rate floating agreement data is used first, all historical data in the past t years are selected as samples, and the exhaustive method of historical data is used to fit and establish the function model y=f(x;A) of the customer's total deposit interval and the interest rate floating interval, where x is the independent variable equal to the customer's total deposit interval, y is equal to the customer's interest rate floating ratio interval, A={α1,α2,α3...}, and A is an exogenous variable that affects the interest rate floating, including customer qualifications, deposit term, and location.

[0159] Among them, α1 is the customer qualification, and whether the customer is a high-quality customer can be judged through the customer classification system established above. α1=1 indicates a high-quality customer. α1=0 indicates other types of customers.

[0160] α2 is the deposit term when the customer makes an interest rate increase, which can be obtained based on the unit's regular deposit certificate.

[0161] α3 is the region where the company is located. According to the unit time deposit confirmation form, the number of deposits with floating interest rates is counted by region, and the regions are divided into high-frequency trading regions and low-frequency trading regions. α3=1 indicates a high-frequency trading region, and α3=0 indicates a low-frequency trading region.

[0162] The above function model y=f(x; A) is specifically expanded as follows:

[0163] y=β0+β1×x+β2×α1+β2×α2×x+β3×α3

[0164] With the deepening of interest rate marketization, interest rate changes are becoming more and more flexible. Therefore, the closer the interest rate increase agreement is signed, the stronger its reference is. Therefore, samples from the past t years are selected to fit the parameters β of each variable, and the corresponding relationship between the total deposit range and the interest rate increase ratio range is obtained through machine learning algorithms. t can be set according to actual conditions, and this application does not impose any restrictions on this.

[0165] Step 13: Split the total deposit interval obtained above into the daily average deposit base + the interval corresponding to the deposit to be increased this time. The customer's daily average deposit base is obtained by adding up the daily average deposits of all accounts opened by the customer last year. The daily average deposit of an account is generally calculated on an annual basis. If the account opening time is less than one year, it is calculated by averaging the number of days the account was opened last year.

[0166] Step 14: According to the above processing, the deposit growth range and interest rate increase range corresponding to different daily average deposit bases are obtained, and the interest rate increase grading model is established. In order to control the deposit cost, the maximum increase ratio λ is set for the excess part. The results of the historical data of the past t years after the above processing can be shown in the above Table 1.

[0167] Step 15: Calculate the deposit ratio based on the customer's new deposit amount, the total balance of the customer's account and the average daily deposit base, and find the corresponding deposit range. The specific algorithm is: deposit ratio = (new deposits + total balance of all customer accounts on that day - average daily deposit base) / average daily deposit base * 100%.

[0168] Assume that the average daily deposit base of this customer last year is x m Yuan, the current total account balance is r yuan, and the above algorithm is used to obtain the deposit ratio p:

[0169]

[0170] Step 16: Based on the deposit corresponding range found according to the deposit ratio calculated above and the customer's average daily deposit base, use the interest rate increase tiering model to find the corresponding interest rate increase range, and calculate the interest rate increase ratio of this deposit according to the following algorithm: interest rate increase ratio = [deposit ratio - deposit corresponding range (lower limit)] × [interest rate increase ratio (upper limit) - interest rate increase ratio (lower limit)] / [deposit corresponding range (upper limit) - deposit corresponding range (lower limit)] + interest rate increase ratio (lower limit).

[0171] According to the deposit ratio calculated above and the above algorithm, the current interest rate increase ratio pi is calculated:

[0172]

[0173] Step 17: Determine the final interest rate for the customer's deposit by multiplying the listed interest rate of the customer's deposit product by (1 + the current interest rate increase ratio pi). Complete the customer's deposit business processing process based on the final interest rate.

[0174] Scenario 2: A new customer plans to deposit d2 yuan in a three-month corporate time deposit, and the deposit exceeds the maximum deposit range.

[0175] Step 21: Since the customer is a new customer, a new customer refers to a customer who has not made a regular deposit and is not included in submodule 2, it is necessary to extract the various characteristic values ​​of the customer according to the method of sending customer classification, so as to calculate the comprehensive score of the customer and determine whether to include it in submodule 2. If the customer's comprehensive score does not meet the standard in submodule 2, it cannot enter the automatic interest rate floating stage. If the customer's comprehensive score can be included in submodule 2, enter Step 2.

[0176] Steps 22 to 25 are similar to the above steps 12 to 15. Assuming that the historical data of the past t years is used to construct the interest rate floating tier model, the average daily deposit x of a customer last year is calculated. m Yuan, the current total current account balance is r yuan. If the above algorithm is used to obtain the customer's deposit ratio: It is determined that the deposit ratio exceeds the maximum deposit range, and the deposit interest rate this time will increase according to the maximum floating ratio λ set by the model.

[0177] Step 26: The listed interest rate of the customer's deposit product is multiplied by (1+λ) to determine the final execution interest rate of the customer's deposit. The customer's deposit business processing process is completed based on the final execution interest rate.

[0178] Scenario 3: An existing customer plans to deposit d3 yuan in a two-year fixed deposit, and the deposit interval after the deposit is negative.

[0179] Steps 31 to 35 are consistent with scenario 1. Assume that the interest rate floating tier model is constructed using the historical data of the past t years. The average daily deposit x of a customer last year is calculated. m Yuan, the current total account balance is r yuan. If the above algorithm is used to obtain the customer's deposit ratio: Then the current interest rate increase ratio is determined to be p i =y n-1 , which is the lower limit of interest rate floating.

[0180] Step 36: Subtract the listed interest rate of the customer's deposit product × (1 + y n-1 ) is determined as the final interest rate for the customer's deposit. The customer's deposit business processing process is completed based on the final interest rate.

[0181] From the above description, it can be seen that the customer classification method provided by this application example can be applied to the calculation of interest rate increase discounts for large corporate customers. First, a corporate customer classifier is constructed through the various types of information stored for corporate customers, and the classifier is used to quickly and effectively screen out high-quality corporate customers. For such high-quality customers, based on the existing manual assessment of interest rate increases, based on the stock data of interest rate increase agreements, an interest rate increase classification model is processed, and an accurate algorithm for calculating the interest rate increase ratio is established in combination with the customer's new deposit number, account balance, etc., to obtain an accurate interest rate increase ratio, completely eliminating manual processing, and providing customers with more flexible, scientific, and efficient interest rate preferential services when making self-service deposits.

[0182] Figure 4 This is a logical diagram of a customer classification method in an example of the present application, wherein the functions implemented by the combination of the customer classification device and the interest rate calculation device can be equivalent to the functions implemented by the above-mentioned customer classification system.

[0183] From the software level, in order to improve the efficiency and reliability of customer classification, and thus improve the reliability of subsequent deposit business processing, the present application provides an embodiment of a customer classification system for implementing all or part of the contents of the customer classification method, see Figure 5 , the customer classification system specifically includes the following contents:

[0184] Acquisition module 01, used to acquire the institutional characteristic data of the financial institution and the characteristic data of the target characteristic type corresponding to the customer to be processed;

[0185] Determination module 02 is used to determine the classification result of the customer to be processed according to the institutional characteristic data of the financial institution, the characteristic data of the target characteristic type and the preset classification model; the target characteristic type is obtained by pre-applying the first characteristic screening method or the second characteristic screening method; the preset classification model is obtained by pre-training the fuzzy clustering algorithm based on the characteristic data set; the application of the first characteristic screening method includes: applying the original characteristic data groups corresponding to each of the plurality of characteristic types to determine the correlation coefficient corresponding to each characteristic type; performing characteristic screening on each characteristic type based on the correlation coefficient to obtain a first intermediate characteristic type; applying the original characteristic data groups corresponding to the first intermediate characteristic type to determine the characteristic contribution of the first intermediate characteristic type; performing characteristic screening on the first intermediate characteristic type based on the characteristic contribution to obtain the target characteristic type; the application of the second characteristic screening method includes: applying the original characteristic data groups corresponding to each characteristic type to determine the correlation coefficient and characteristic contribution of each characteristic type; performing characteristic screening on each characteristic type based on the correlation coefficient to obtain the first intermediate characteristic type; performing characteristic screening on each characteristic type based on the characteristic contribution to obtain the second intermediate characteristic type; and determining the characteristic type belonging to both the first intermediate characteristic type and the second intermediate characteristic type as the target characteristic type.

[0186] In one embodiment, the customer classification system further includes:

[0187] The data set acquisition module is used to acquire an original feature data set, wherein the original feature data set includes: a plurality of original samples, each of which includes: original feature data corresponding to a plurality of feature types; the original feature data set is divided into a plurality of groups of original feature data groups, wherein the feature types corresponding to the original feature data in the same group of original feature data groups are the same, and the feature types corresponding to the original feature data in different groups of original feature data groups are different;

[0188] The screening module is used to apply the first feature screening method or the second feature screening method based on the original feature data groups corresponding to each feature type, to screen out the target feature type from each feature type.

[0189] In one embodiment, the application of the original feature data groups corresponding to each feature type to determine the correlation coefficient and feature contribution of each feature type includes:

[0190] Obtain the actual classification results corresponding to each original sample;

[0191] The classification algorithm is trained using each original sample and its corresponding actual classification results to obtain a preset initial classification model;

[0192] Input each original feature data in the same original feature data group into a preset initial classification model to obtain a prediction classification result corresponding to the original feature data;

[0193] Determine the actual classification result corresponding to the original sample to which the original feature data belongs as the actual classification result corresponding to the original feature data;

[0194] According to the actual classification results and predicted classification results corresponding to each original feature data in the same original feature data group, the feature contribution of the feature type corresponding to the original feature data group is determined; and the correlation coefficient of each feature type is determined by applying the original feature data groups corresponding to each feature type.

[0195] In one embodiment, the customer classification system further includes:

[0196] A sample acquisition module is used to acquire a feature data set, the feature data set comprising: a plurality of samples, each sample comprising: historical institution feature data and historical feature data of a target feature type;

[0197] The training module is used to train the fuzzy clustering algorithm using the feature data set to obtain the classification model.

[0198] In one embodiment, the customer classification system further includes:

[0199] A receiving module, used for receiving the deposit request of the customer to be processed;

[0200] Correspondingly, the customer classification system further includes:

[0201] A module for determining an interest rate increase ratio is used to determine the current interest rate increase ratio corresponding to the customer to be processed according to the deposit request if the classification result is a high-quality customer;

[0202] A module for determining a final execution ratio is used to determine the final execution ratio of the customer to be processed according to the pre-acquired listing ratio and the current interest rate increase ratio;

[0203] A processing module is used to complete the deposit business processing process corresponding to the deposit request according to the final execution interest rate.

[0204] In one embodiment, the module for determining the interest rate increase ratio includes:

[0205] A region type determination unit, configured to determine a total deposit interval and a region type corresponding to the customer to be processed according to the pre-acquired average daily deposit base corresponding to the customer to be processed, the deposit amount and the location information in the deposit request;

[0206] A determination ratio unit is used to determine the interest rate increase ratio corresponding to the customer to be processed based on a preset interest rate increase ratio model, the pre-acquired average daily deposit base, the classification result, the total deposit range and the type of region; wherein the preset interest rate increase ratio model is obtained by pre-training the regression model using batch samples and their respective corresponding real interest rate increase ratios, and each sample includes: the average daily deposit base of historical customers, the classification result, the deposit range and the type of region.

[0207] In one embodiment, the module for determining the interest rate increase ratio includes:

[0208] The first determining unit is used to determine the deposit ratio of the customer to be processed according to the pre-acquired total account balance corresponding to the customer to be processed, the average daily deposit base and the deposit amount in the deposit request.

[0209] The second determination unit is used to determine the deposit corresponding interval and interest rate floating ratio interval corresponding to the customer to be processed according to a preset corresponding relationship table, the pre-acquired average daily deposit base and the deposit ratio, and the preset corresponding relationship table includes: the correspondence between multiple average daily deposit base intervals, deposit corresponding intervals and interest rate floating intervals.

[0210] The third determining unit is used to determine the current interest rate increase ratio corresponding to the customer to be processed according to the deposit ratio, the deposit corresponding interval and the interest rate increase ratio interval.

[0211] The embodiment of the customer classification system provided in this specification can be specifically used to execute the processing flow of the embodiment of the above-mentioned customer classification method. Its functions will not be described in detail here, and reference can be made to the detailed description of the embodiment of the above-mentioned customer classification method.

[0212] From the hardware level, in order to improve the efficiency and reliability of customer classification, and thus improve the reliability of subsequent deposit business processing, the present application provides an embodiment of an electronic device for implementing all or part of the contents of the customer classification method. The electronic device specifically includes the following contents:

[0213] A processor, a memory, a communications interface and a bus; wherein the processor, the memory and the communications interface communicate with each other through the bus; the communications interface is used to realize information transmission between the customer classification system and related devices such as user terminals; the electronic device can be a desktop computer, a tablet computer and a mobile terminal, etc., but the present embodiment is not limited thereto. In the present embodiment, the electronic device can be implemented with reference to the embodiment for realizing the customer classification method and the embodiment for realizing the customer classification system, and the contents thereof are incorporated herein, and the repeated parts are not repeated.

[0214] Figure 6 FIG. 9 is a schematic block diagram of the system structure of the electronic device 9600 according to an embodiment of the present application. Figure 6 As shown, the electronic device 9600 may include a central processor 9100 and a memory 9140; the memory 9140 is coupled to the central processor 9100. It is worth noting that Figure 6 is exemplary; other types of structures may also be used to supplement or replace this structure to implement telecommunication functions or other functions.

[0215] In one or more embodiments of the present application, the customer classification function may be integrated into the central processor 9100. The central processor 9100 may be configured to perform the following control:

[0216] Step 100: Acquire the institutional characteristic data of the financial institution and the characteristic data of the target characteristic type corresponding to the customer to be processed;

[0217] Step 200: Determine the classification result of the customer to be processed according to the institutional characteristic data of the financial institution, the characteristic data of the target characteristic type and a preset classification model;

[0218] Step 300: The target feature type is obtained by pre-applying the first feature screening method or the second feature screening method; the preset classification model is obtained by pre-training the fuzzy clustering algorithm based on the feature data set;

[0219] The first feature screening method includes: applying the original feature data groups corresponding to the plurality of feature types to determine the correlation coefficients corresponding to the various feature types; performing feature screening on the various feature types based on the correlation coefficients to obtain the first intermediate feature type; applying the original feature data groups corresponding to the first intermediate feature type to determine the feature contribution of the first intermediate feature type; performing feature screening on the first intermediate feature type based on the feature contribution to obtain the target feature type;

[0220] The second feature screening method includes: applying the original feature data groups corresponding to each feature type to determine the correlation coefficient and feature contribution of each feature type; performing feature screening on each feature type based on the correlation coefficient to obtain a first intermediate feature type; performing feature screening on each feature type based on the feature contribution to obtain a second intermediate feature type; and determining a feature type that belongs to both the first intermediate feature type and the second intermediate feature type as the target feature type.

[0221] It can be seen from the above description that the electronic device provided by the embodiment of the present application can improve the efficiency and reliability of customer classification, thereby improving the reliability of subsequent deposit business processing.

[0222] In another embodiment, the customer classification system may be configured separately from the central processor 9100. For example, the customer classification system may be configured as a chip connected to the central processor 9100, and the customer classification function may be implemented under the control of the central processor.

[0223] like Figure 6 As shown, the electronic device 9600 may also include: a communication module 9110, an input unit 9120, an audio processor 9130, a display 9160, and a power supply 9170. It is worth noting that the electronic device 9600 does not necessarily have to include Figure 6 In addition, the electronic device 9600 may also include Figure 6 For components not shown, reference may be made to the prior art.

[0224] like Figure 6 As shown, the central processing unit 9100 is sometimes also referred to as a controller or an operation control, and may include a microprocessor or other processor device and / or logic device. The central processing unit 9100 receives input and controls the operation of various components of the electronic device 9600.

[0225] The memory 9140 may be, for example, one or more of a cache, a flash memory, a hard drive, a removable medium, a volatile memory, a non-volatile memory or other suitable devices. The above-mentioned information related to the failure may be stored, and a program for executing the relevant information may also be stored. The CPU 9100 may execute the program stored in the memory 9140 to implement information storage or processing, etc.

[0226] The input unit 9120 provides input to the central processing unit 9100. The input unit 9120 is, for example, a key or a touch input device. The power supply 9170 is used to provide power to the electronic device 9600. The display 9160 is used to display display objects such as images and texts. The display may be, for example, an LCD display, but is not limited thereto.

[0227] The memory 9140 may be a solid-state memory, such as a read-only memory (ROM), a random access memory (RAM), a SIM card, etc. It may also be a memory that saves information even when the power is off, can be selectively erased, and is provided with more data, examples of which are sometimes referred to as EPROMs, etc. The memory 9140 may also be some other type of device. The memory 9140 includes a buffer memory 9141 (sometimes referred to as a buffer). The memory 9140 may include an application / function storage unit 9142, which is used to store application programs and function programs or processes for executing the operation of the electronic device 9600 through the central processor 9100.

[0228] The memory 9140 may also include a data storage unit 9143 for storing data, such as contacts, digital data, pictures, sounds, and / or any other data used by the electronic device. The driver storage unit 9144 of the memory 9140 may include various drivers for communication functions of the electronic device and / or for executing other functions of the electronic device (such as messaging applications, address book applications, etc.).

[0229] The communication module 9110 is a transmitter / receiver 9110 that sends and receives signals via an antenna 9111. The communication module (transmitter / receiver) 9110 is coupled to the central processor 9100 to provide input signals and receive output signals, which may be the same as the case of a conventional mobile communication terminal.

[0230] Based on different communication technologies, multiple communication modules 9110 may be provided in the same electronic device, such as a cellular network module, a Bluetooth module and / or a wireless LAN module, etc. The communication module (transmitter / receiver) 9110 is also coupled to a speaker 9131 and a microphone 9132 via an audio processor 9130 to provide an audio output via the speaker 9131 and receive an audio input from the microphone 9132, thereby realizing a common telecommunication function. The audio processor 9130 may include any suitable buffer, decoder, amplifier, etc. In addition, the audio processor 9130 is also coupled to the central processor 9100, so that recording can be performed on the local machine through the microphone 9132, and the sound stored on the local machine can be played through the speaker 9131.

[0231] It can be seen from the above description that the electronic device provided by the embodiments of the present application can improve the efficiency and reliability of customer classification, thereby improving the reliability of subsequent deposit business processing.

[0232] The embodiments of the present application also provide a computer-readable storage medium capable of implementing all the steps in the customer classification method in the above embodiments. The computer-readable storage medium stores a computer program. When the computer program is executed by a processor, all the steps in the customer classification method in the above embodiments are implemented. For example, when the processor executes the computer program, the following steps are implemented:

[0233] Step 100: Acquire the institutional characteristic data of the financial institution and the characteristic data of the target characteristic type corresponding to the customer to be processed;

[0234] Step 200: Determine the classification result of the customer to be processed according to the institutional characteristic data of the financial institution, the characteristic data of the target characteristic type and a preset classification model;

[0235] Step 300: The target feature type is obtained by pre-applying the first feature screening method or the second feature screening method; the preset classification model is obtained by pre-training the fuzzy clustering algorithm based on the feature data set;

[0236] The first feature screening method includes: applying the original feature data groups corresponding to the plurality of feature types to determine the correlation coefficients corresponding to the various feature types; performing feature screening on the various feature types based on the correlation coefficients to obtain the first intermediate feature type; applying the original feature data groups corresponding to the first intermediate feature type to determine the feature contribution of the first intermediate feature type; performing feature screening on the first intermediate feature type based on the feature contribution to obtain the target feature type;

[0237] The second feature screening method includes: applying the original feature data groups corresponding to each feature type to determine the correlation coefficient and feature contribution of each feature type; performing feature screening on each feature type based on the correlation coefficient to obtain a first intermediate feature type; performing feature screening on each feature type based on the feature contribution to obtain a second intermediate feature type; and determining a feature type that belongs to both the first intermediate feature type and the second intermediate feature type as the target feature type.

[0238] It can be seen from the above description that the computer-readable storage medium provided in the embodiment of the present application can improve the efficiency and reliability of customer classification, thereby improving the reliability of subsequent deposit business processing.

[0239] In this application, each embodiment of the above method is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. For relevant parts, refer to the partial description of the method embodiment.

[0240] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented in one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that include computer-usable program code.

[0241] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0242] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0243] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0244] Specific embodiments are used in this application to illustrate the principles and implementation methods of this application. The description of the above embodiments is only used to help understand the method and core idea of ​​this application. At the same time, for those skilled in the art, according to the idea of ​​this application, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as a limitation on this application.

Claims

1. A customer classification method, characterized in that: include: Obtaining the institutional characteristic data of the financial institution and the characteristic data of the target characteristic type corresponding to the customer to be processed; Determining the classification result of the customer to be processed according to the institutional characteristic data of the financial institution, the characteristic data of the target characteristic type and a preset classification model; The target feature type is obtained by pre-applying the first feature screening method or the second feature screening method; The preset classification model is obtained by pre-training the fuzzy clustering algorithm based on the feature data set; The first feature screening method includes: applying the original feature data groups corresponding to the plurality of feature types to determine the correlation coefficients corresponding to the various feature types; performing feature screening on the various feature types based on the correlation coefficients to obtain the first intermediate feature type; applying the original feature data groups corresponding to the first intermediate feature type to determine the feature contribution of the first intermediate feature type; performing feature screening on the first intermediate feature type based on the feature contribution to obtain the target feature type; The second feature screening method includes: applying the original feature data groups corresponding to each feature type to determine the correlation coefficient and feature contribution of each feature type; performing feature screening on each feature type based on the correlation coefficient to obtain a first intermediate feature type; performing feature screening on each feature type based on the feature contribution to obtain a second intermediate feature type; and determining a feature type that belongs to both the first intermediate feature type and the second intermediate feature type as the target feature type.

2. The customer classification method according to claim 1, characterized in that: Before obtaining the institutional characteristic data of the financial institution and the characteristic data of the target characteristic type corresponding to the customer to be processed, the method further includes: Acquire an original feature data set, the original feature data set comprising: a plurality of original samples, each original sample comprising: original feature data corresponding to a plurality of feature types; divide the original feature data set into a plurality of groups of original feature data groups, the original feature data in the same group of original feature data groups correspond to the same feature type, and the feature types corresponding to the original feature data in different groups of original feature data groups are different; Based on the original feature data groups corresponding to each of the feature types, the first feature screening method or the second feature screening method is applied to screen out the target feature type from each of the feature types.

3. The customer classification method according to claim 2, characterized in that: The method of applying the original feature data groups corresponding to each feature type to determine the correlation coefficient and feature contribution of each feature type includes: Obtain the actual classification results corresponding to each original sample; The classification algorithm is trained using each original sample and its corresponding actual classification results to obtain a preset initial classification model; Input each original feature data in the same original feature data group into a preset initial classification model to obtain a prediction classification result corresponding to the original feature data; Determine the actual classification result corresponding to the original sample to which the original feature data belongs as the actual classification result corresponding to the original feature data; According to the actual classification results and predicted classification results corresponding to each original feature data in the same original feature data group, the feature contribution of the feature type corresponding to the original feature data group is determined; and the correlation coefficient of each feature type is determined by applying the original feature data groups corresponding to each feature type.

4. The customer classification method according to claim 1, characterized in that: Before determining the classification result of the customer to be processed according to the institutional characteristic data of the financial institution, the characteristic data of the target characteristic type and the preset classification model, the method further includes: Acquire a feature data set, the feature data set comprising: a plurality of samples, each sample comprising: historical institution feature data and historical feature data of a target feature type; The feature data set is used to train a fuzzy clustering algorithm to obtain the classification model.

5. The customer classification method according to claim 1, characterized in that: Before obtaining the institutional characteristic data of the financial institution and the characteristic data of the target characteristic type corresponding to the customer to be processed, the method further includes: receiving a deposit request from the pending customer; Correspondingly, after determining the classification result of the customer to be processed according to the institutional characteristic data of the financial institution, the characteristic data of the target characteristic type and the preset classification model, the method further includes: If the classification result is a high-quality customer, then determine the current interest rate increase ratio corresponding to the customer to be processed according to the deposit request; Determine the final execution interest rate of the customer to be processed based on the pre-obtained listed interest rate and the current interest rate increase ratio; According to the final execution interest rate, the deposit business processing process corresponding to the deposit request is completed.

6. The customer classification method according to claim 5, characterized in that: Determining the current interest rate increase ratio corresponding to the pending customer according to the deposit request includes: Determine the total deposit interval and the type of region corresponding to the customer to be processed according to the pre-acquired average daily deposit base corresponding to the customer to be processed, the deposit amount and the location information in the deposit request; Determine the interest rate increase ratio corresponding to the customer to be processed according to the preset interest rate increase ratio model, the pre-acquired average daily deposit base, the classification result, the total deposit range and the type of region; Among them, the preset interest rate increase ratio model is obtained by pre-training the regression model using batch samples and their corresponding real interest rate increase ratios, and each sample includes: the historical customer's average daily deposit base, classification results, deposit range and region type.

7. The customer classification method according to claim 5, characterized in that: Determining the current interest rate increase ratio corresponding to the pending customer according to the deposit request includes: Determine the deposit ratio of the customer to be processed according to the pre-acquired total account balance, the daily average deposit base and the deposit amount in the deposit request corresponding to the customer to be processed; Determine the deposit corresponding interval and interest rate increase ratio interval corresponding to the customer to be processed according to the preset corresponding relationship table, the pre-acquired daily average deposit base and the deposit ratio, wherein the preset corresponding relationship table includes: the corresponding relationship between multiple daily average deposit base intervals, deposit corresponding intervals and interest rate increase intervals; The current interest rate increase ratio corresponding to the customer to be processed is determined according to the deposit ratio, the deposit corresponding range and the interest rate increase ratio range.

8. A customer classification system, characterized in that: include: An acquisition module, used to acquire the institutional characteristic data of the financial institution and the characteristic data of the target characteristic type corresponding to the customer to be processed; A determination module, used to determine the classification result of the customer to be processed according to the institutional characteristic data of the financial institution, the characteristic data of the target characteristic type and a preset classification model; The target feature type is obtained by pre-applying the first feature screening method or the second feature screening method; The preset classification model is obtained by pre-training the fuzzy clustering algorithm based on the feature data set; The first feature screening method includes: applying the original feature data groups corresponding to the plurality of feature types to determine the correlation coefficients corresponding to the various feature types; performing feature screening on the various feature types based on the correlation coefficients to obtain the first intermediate feature type; applying the original feature data groups corresponding to the first intermediate feature type to determine the feature contribution of the first intermediate feature type; performing feature screening on the first intermediate feature type based on the feature contribution to obtain the target feature type; The second feature screening method includes: applying the original feature data groups corresponding to each feature type to determine the correlation coefficient and feature contribution of each feature type; performing feature screening on each feature type based on the correlation coefficient to obtain a first intermediate feature type; performing feature screening on each feature type based on the feature contribution to obtain a second intermediate feature type; and determining a feature type that belongs to both the first intermediate feature type and the second intermediate feature type as the target feature type.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the program, the customer classification method according to any one of claims 1 to 7 is implemented.

10. A computer-readable storage medium having computer instructions stored thereon, characterized in that: When the instructions are executed by a processor, the customer classification method described in any one of claims 1 to 7 is implemented.

Images