Attribution analysis method and system for overdue index change

Abstract

The invention provides an attribution analysis method and system for overdue index change, and belongs to the technical field of financial information, and the method specifically comprises the steps: employing the deviation condition of the distribution data of a change index characteristic in different intervals and other unit time periods in an index abnormal time period, determining potential association index characteristics in the change index characteristics, and determining the change index characteristics based on the change condition of distribution data and the change condition of credit indexes of detail index characteristics in different intervals in unit time periods with different potential association index characteristics in a specified time period; the correlation coefficients of different potential correlation index features and the correlation index features are determined, and the credit index change attribution analysis is performed according to the sorting result of the correlation coefficients of the correlation index features, so that the processing accuracy of the attribution analysis is improved.

Description

Technical Field

[0001] The present invention belongs to the field of financial information technology, and in particular relates to an attribution analysis method and system for overdue indicator changes. Background Art

[0002] In the process of carrying out credit business, it is necessary to focus on credit indicators in multiple dimensions, such as collection rate, migration rate, collection rate, bad debt rate, overdue rate, etc. Generally, if there is an abnormality in a certain indicator and it cannot be handled in time, it will inevitably cause losses to the credit institution. Therefore, how to quickly realize the attribution analysis of indicator changes and timely control has become a technical problem that needs to be solved urgently.

[0003] In order to realize the attribution analysis of the indicator changes of credit indicators, the invention patent application CN202410258403.X "A Credit Risk Abnormal Inspection Attribution Warning Method and System" determines the credit characteristics of interest by analyzing the correlation between the risk indicators and the indicator changes of the credit characteristics, so that the risk control strategy can be optimized according to the correlation between the indicator changes of the credit characteristics of interest, reducing the risk of credit application processing, but there are the following technical problems: When credit indicators change, it is often caused by changes in credit characteristics in multiple dimensions. The credit characteristics of any single dimension have not changed significantly. Therefore, if attribution analysis cannot be performed based on the ranking results of credit characteristics, it may lead to focusing only on one dimension and failing to accurately implement attribution analysis of the credit characteristics.

[0004] In response to the above technical problems, the present invention provides an attribution analysis method and system for changes in overdue indicators. Summary of the invention

[0005] To achieve the purpose of the present invention, the present invention adopts the following technical solutions: According to one aspect of the present invention, there is provided an attribution analysis method for changes in overdue indicators.

[0006] An attribution analysis method for changes in overdue indicators, specifically comprising: S1 determines the abnormal period of the indicator based on the change data of the credit indicator of the credit business within the specified period, and determines the credit indicator of concern among the credit indicators in combination with the fluctuation of the credit characteristics between different unit periods; S2 divides the credit application users of the credit business into multiple intervals based on the characteristic data under the detailed indicator characteristics, and determines the change indicator characteristics in the detailed indicator characteristics according to the change of the distribution data of the detailed indicator characteristics in different unit time periods in the specified time period in different intervals; S3 determines the potential correlation indicator features in the change indicator features by using the deviation of the distribution data of different intervals in the indicator abnormal period and other unit time periods; S4 determines the correlation coefficients and correlation indicator characteristics of different potential correlation indicator characteristics based on the changes in the distribution data of the detailed indicator characteristics in different unit time periods of the potential correlation indicator characteristics in the specified time period and the changes in the credit indicators, and performs attribution analysis of the changes in the credit indicators based on the sorting results of the correlation coefficients of the correlation indicator characteristics.

[0007] A further technical solution is that the designated period is determined according to the number of credit application users of the credit business, wherein the greater the number of credit users of the credit business, the shorter the designated period is.

[0008] A further technical solution is that the credit indicators include collection rate, migration rate, collection recovery rate, non-performing loan rate and overdue rate.

[0009] A further technical solution is that the abnormal indicator period is a unit period in which the credit indicator is not within a preset indicator range.

[0010] A further technical solution is that the method for determining the characteristics of the correlation index is: Based on the changes in the distribution data of the detailed indicator characteristics in different unit time periods with different characteristics of the potential correlation indicator and the changes in the credit indicators, the change correlation coefficients in different unit time periods are determined; Based on the change correlation coefficients in different unit time periods, the correlation coefficients of the potential correlation indicator features are determined, and the correlation indicator features in the potential correlation indicator features are determined based on the correlation coefficients.

[0011] A further technical solution is that the correlation indicator feature is a potential correlation indicator feature whose correlation coefficient is greater than a preset correlation coefficient threshold.

[0012] In a second aspect, the present invention provides a computer system comprising: a memory and a processor connected in communication, and a computer program stored in the memory and capable of running on the processor, characterized in that: when the processor runs the computer program, it executes the above-mentioned attribution analysis method for changes in overdue indicators.

[0013] In a third aspect, the present invention provides an attribution analysis model for overdue indicator changes, using the above-mentioned attribution analysis method for overdue indicator changes, specifically including: Focus on the target screening module, change feature screening module, correlation feature screening module, and attribution analysis processing module; The focus target screening module is responsible for determining the abnormal indicator period based on the change data of the credit indicator of the credit business within the specified period, and determining the focus credit indicator among the credit indicators in combination with the fluctuation of the credit characteristics between different unit time periods; The change feature screening module is responsible for dividing the credit application users of the credit business into multiple intervals based on the feature data under the detailed indicator feature, and determining the change indicator feature in the detailed indicator feature according to the change of the distribution data of the detailed indicator feature in different unit time periods in different intervals in the specified time period; The associated feature screening module is responsible for determining potential associated indicator features in the change indicator features by using the distribution data of different intervals of the change indicator features within the indicator abnormal period and the deviations within other unit time periods; The attribution analysis processing module is responsible for determining the correlation coefficients and correlation indicator characteristics of different potential correlation indicator characteristics based on the changes in the distribution data of the detailed indicator characteristics in different unit time periods of the potential correlation indicator characteristics in the specified time period and the changes in the credit indicators, and performing attribution analysis of the changes in credit indicators based on the sorting results of the correlation coefficients of the correlation indicator characteristics.

[0014] The beneficial effects of the present invention are: 1. Determine the credit indicators of concern among the credit indicators based on the fluctuations of credit characteristics during periods of indicator abnormality and between different unit time periods. This enables the screening of credit indicators of concern from two perspectives: the fluctuation of credit characteristics within a specified period and the abnormal periods of credit characteristics within a specified period. It also lays the foundation for conducting targeted attribution analysis on credit indicators with higher volatility and higher degree of indicator abnormality.

[0015] 2. The attribution analysis of changes in credit indicators is performed based on the ranking results of the correlation coefficients of the related indicator characteristics, which avoids the technical problem that the accuracy of the attribution analysis results cannot meet the requirements due to the single consideration of a certain related indicator characteristic. It realizes the comprehensive consideration of multiple related indicator characteristics from the perspective of the ranking results, ensures the accuracy of the attribution analysis processing, and also lays the foundation for further generating differentiated intervention methods to avoid losses of credit institutions.

[0016] Other features and advantages will be described in the following description. The objects and other advantages of the present invention are realized and obtained by the structures particularly pointed out in the description and drawings.

[0017] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, preferred embodiments are given below and described in detail with reference to the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] The above and other features and advantages of the present invention will become more apparent by describing in detail exemplary embodiments thereof with reference to the attached drawings.

[0019] Figure 1 It is a flow chart of an attribution analysis method for changes in overdue indicators; Figure 2 It is a flow chart of a method for determining a credit indicator of interest in a credit indicator; Figure 3 It is a flow chart of a method for determining a variable indicator feature in a detailed indicator feature; Figure 4 is a flow chart of a method for determining characteristics of an associated indicator; Figure 5 It is a framework diagram of a computer system. DETAILED DESCRIPTION

[0020] 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 specification will be clearly and completely described below in conjunction with the drawings in the embodiments of this specification. Obviously, the described embodiments are only part of the embodiments of this specification, not all of the embodiments. Based on the embodiments of this specification, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of this specification.

[0021] When credit indicators change, it is often caused by changes in credit characteristics in multiple dimensions. The credit characteristics in any single dimension have not changed significantly. Therefore, it is necessary to conduct attribution analysis based on the ranking results of credit characteristics to realize the attribution analysis of the occurrence of credit characteristics.

[0022] Abnormal indicator period: a unit period when the credit indicator is not within the preset indicator range.

[0023] Focus on credit indicators: Based on the fluctuation of credit characteristics between different unit time periods, determine the change in credit characteristics between different adjacent unit time periods, and use the unit time period with a change greater than the preset change as the changing unit time period. The total number of time periods is determined based on the number of abnormal indicator time periods and changing unit time periods, and use the credit indicators with a total number of time periods within the preset number range as the focus credit indicators among the credit indicators.

[0024] Changing indicator characteristics: determine the quantity proportion in different intervals within the unit time period based on the distribution data of the detailed indicator characteristics in different intervals within the unit time period, use the quantity proportion to determine the deviation of the quantity proportion in different intervals between adjacent unit time periods, and use the average value of the deviation of the quantity proportion in different intervals to determine the characteristic variation coefficient with adjacent unit time periods, determine the comprehensive variation coefficient based on the average value of the characteristic variation coefficients of different unit time periods, and use the detailed indicator characteristics with a comprehensive variation coefficient greater than 0.6 as changing indicator characteristics.

[0025] Potentially correlated indicator characteristics: determine the proportion of the number of different intervals in the unit time period within the indicator abnormal period of the changing indicator characteristics, and determine the interval with deviations from the quantity proportions in other unit time periods based on the deviations from other unit time periods, and use it as the distribution deviation interval, and determine the deviation period proportion factor of different indicator abnormal periods by the quantity proportion of other unit time periods with the distribution deviation interval, and use the average value of the deviation period proportion factors of different indicator abnormal periods to determine the characteristic distribution deviation coefficient of the changing indicator characteristics, and determine the changing indicator characteristics whose characteristic distribution deviation coefficient is greater than 0.5 as potential correlated indicator characteristics. Correlation indicator characteristics: Based on the changes in the distribution data of detailed indicator characteristics in different unit time periods with potential correlation indicator characteristics and the changes in credit indicators, the change correlation coefficients in different unit time periods are determined. Based on the average value of the change correlation coefficients in different unit time periods, the correlation coefficient of the potential correlation indicator characteristics is determined, and the potential correlation indicator characteristics with a correlation coefficient greater than 0.6 are taken as correlation indicator characteristics. Embodiment 1 To solve the above problem, according to one aspect of the present invention, Figure 1 As shown, an attribution analysis method for changes in overdue indicators is provided, which specifically includes: S1 determines the abnormal period of the indicator based on the change data of the credit indicator of the credit business within the specified period, and determines the credit indicator of concern among the credit indicators in combination with the fluctuation of the credit characteristics between different unit periods; S2 divides the credit application users of the credit business into multiple intervals based on the characteristic data under the detailed indicator characteristics, and determines the change indicator characteristics in the detailed indicator characteristics according to the change of the distribution data of the detailed indicator characteristics in different unit time periods in the specified time period in different intervals; S3 determines the potential correlation indicator features in the change indicator features by using the deviation of the distribution data of different intervals in the indicator abnormal period and other unit time periods; S4 determines the correlation coefficients and correlation indicator characteristics of different potential correlation indicator characteristics based on the changes in the distribution data of the detailed indicator characteristics in different unit time periods of the potential correlation indicator characteristics in the specified time period and the changes in the credit indicators, and performs attribution analysis of the changes in the credit indicators based on the sorting results of the correlation coefficients of the correlation indicator characteristics.

[0026] Furthermore, the designated period is determined according to the number of credit application users of the credit service, wherein the greater the number of credit users of the credit service, the shorter the designated period is.

[0027] Specifically, the credit indicators include collection rate, migration rate, collection recovery rate, non-performing loan rate and overdue rate.

[0028] It can be understood that the abnormal indicator period is a unit period when the credit indicator is not within the preset indicator range.

[0029] Specifically, Figure 2 As shown, the method for determining the credit indicator of interest in the credit indicator is: Obtaining the abnormal time periods of the credit indicator within the specified time period, and determining the abnormal coefficient of the credit indicator by using the proportion of the number of the abnormal time periods within the specified time period; Based on the fluctuation of credit characteristics between different unit time periods, the variation of credit characteristics between different adjacent unit time periods is determined, and the variation unit time period within the unit time period is determined according to the variation, and the index variation coefficient of the credit index is determined according to the proportion of the variation time period within the specified time period; The attention coefficient of the credit indicator is determined based on the indicator variation coefficient and the indicator abnormality coefficient, and the attention credit indicator in the credit indicator is determined through the attention coefficient.

[0030] Furthermore, the credit indicator of concern is a credit indicator whose concern coefficient is greater than a preset coefficient threshold.

[0031] It should also be noted that the attention coefficient is determined based on the average value of the indicator variation coefficient and the indicator abnormality coefficient.

[0032] Optionally, the method for determining the credit indicator of concern in the credit indicator is: Obtain the abnormal time period of the credit indicator within the specified time period; Based on the fluctuation of credit characteristics between different unit time periods, the variation of credit characteristics between different adjacent unit time periods is determined, and the variation unit time period within the unit time period is determined according to the variation; The total number of time periods is determined based on the number of abnormal time periods and unit time periods of the indicator, and the total number of time periods is used to determine the credit indicator of concern among the credit indicators.

[0033] Furthermore, when the total quantity in the time period is not within a preset quantity range, the credit indicator is determined to be a credit indicator of concern.

[0034] Optionally, the method for determining the credit indicator of interest in the credit indicator comprises steps S11-S13, specifically: S11 determines the variation of the credit characteristics between different adjacent unit time periods based on the fluctuation of the credit characteristics between different unit time periods, and determines the variation unit time period within the unit time period according to the variation, and determines the index variation coefficient of the credit index by the proportion of the variation time period in the specified time period and the variation of the credit characteristics between the variation time period and the adjacent unit time periods; S12 obtains the abnormal time period of the credit indicator within the specified time period, and determines the abnormal coefficient of the credit indicator by using the proportion of the abnormal time period within the specified time period and the characteristic quantity of the credit indicator during the abnormal time period; S13 determines the attention coefficient of the credit indicator based on the indicator variation coefficient and the indicator abnormality coefficient, and determines the attention credit indicator in the credit indicator through the attention coefficient.

[0035] Optionally, the above step S11 includes steps S111-S113, which are specifically: S111 determines the variation of the credit characteristics between different adjacent unit time periods based on the fluctuation of the credit characteristics between different unit time periods, and when it is determined according to the variation that there is no variation unit time period within the unit time period, it is determined that the credit indicator does not belong to the credit indicator of interest, and when there is a variation unit time period, proceeds to step S112; S112: when the proportion of the number of the change period in the specified period is greater than the preset proportion, it is determined that the credit indicator belongs to the credit indicator of concern; when the proportion of the number of the change period in the specified period is not greater than the preset proportion, the process proceeds to step S113; S113 determines the index variation coefficient of the credit indicator by the proportion of the number of the change period in the specified time period and the change amount of the credit characteristics between the change period and the adjacent unit time period. When the index variation coefficient of the credit indicator is not within the preset range, proceed to step S12. When the index variation coefficient of the credit indicator is within the preset range, the index variation coefficient of the credit indicator is used to determine whether the credit indicator belongs to the credit indicator of concern.

[0036] It should be noted that the coefficient of variation of the credit indicator is used to determine whether the credit indicator belongs to the credit indicator of concern, specifically: When the index variation coefficient of the credit index is greater than a preset variation coefficient threshold, the credit index is determined to be a credit index of concern; if not, it is directly determined that the credit index does not belong to the credit index of concern.

[0037] Optionally, the above step S12 includes steps S121-S123, specifically: S121 obtains the abnormal time periods of the credit indicator within the specified time period. When the number or the proportion of the abnormal time periods does not meet the requirements, the credit indicator is determined to be a concerned credit indicator. When the number and the proportion of the abnormal time periods meet the requirements, the process proceeds to step S122. S122: when there is an abnormal period of time during which the characteristic amount of the credit characteristic is greater than the preset characteristic amount threshold, the process proceeds to step S123; when there is no abnormal period of time during which the characteristic amount of the credit characteristic is greater than the preset characteristic amount threshold, the credit indicator is determined to be a credit indicator of concern; S123 determines the indicator abnormality coefficient of the credit indicator by utilizing the proportion of the indicator abnormality period in the specified period and the characteristic quantity of the credit indicator in the indicator abnormality period. When the indicator abnormality coefficient of the credit indicator does not meet the requirements, the credit indicator is determined to be a credit indicator of concern. When the indicator abnormality coefficient of the credit indicator meets the requirements, step S13 is entered.

[0038] Furthermore, the credit application users of the credit business are divided into multiple intervals, specifically including: According to the interval in which the characteristic data of the credit applicant is located under the detailed indicator characteristics, the credit applicant is divided into a corresponding interval.

[0039] It should be noted that if Figure 3 As shown, the method for determining the change indicator feature in the detailed indicator feature is: Determine the quantity proportion in different intervals within the unit time period based on the distribution data of the detailed indicator characteristics in different intervals within the unit time period; Determine the deviation interval between adjacent unit time periods by using the quantity proportion, and determine the characteristic variation coefficient between adjacent unit time periods according to the quantity of the deviation interval; The distribution change period in the unit period is determined based on the characteristic change coefficient, and the change index feature in the detailed index feature is determined using the number of the distribution change periods.

[0040] Furthermore, the change indicator feature is a detailed indicator feature in which the number of distribution change time periods is greater than the number of preset time periods.

[0041] It is understandable that the detailed indicator characteristics include the credit applicant's work information, educational information, credit application information, credit expenditure information and overdue information.

[0042] It should also be noted that the method for determining the change indicator feature in the detailed indicator feature is: Determine the quantity proportion in different intervals within the unit time period based on the distribution data of the detailed indicator characteristics in different intervals within the unit time period; Determine the deviation of the quantity proportions in different intervals between adjacent unit time periods by using the quantity proportions, and determine the characteristic variation coefficient between adjacent unit time periods by using the average value of the deviation of the quantity proportions in different intervals; The comprehensive coefficient of variation is determined based on the average values ​​of the characteristic coefficients of variation of different unit time periods, and the variation index characteristics in the detailed index characteristics are determined using the comprehensive coefficient of variation.

[0043] Furthermore, the variation indicator characteristic is a detailed indicator characteristic whose comprehensive variation coefficient does not meet the requirements.

[0044] Optionally, the determination of the change indicator feature in the detailed indicator feature specifically includes the following steps S21-S23: S21 determines the quantity proportions in different intervals within the unit time period based on the distribution data of the detailed indicator characteristics in different intervals within the unit time period; S22: using the quantity proportion to determine the deviation of the quantity proportion in different intervals between adjacent unit time periods, and using the average value of the deviation of the quantity proportion in different intervals and the number of deviation intervals to determine the characteristic variation coefficient between adjacent unit time periods; S23 determines the distribution change period in the unit period based on the characteristic change coefficients of different unit periods, determines the comprehensive change coefficient based on the number of the distribution periods and the characteristic change coefficients of different distribution change periods, and uses the comprehensive change coefficient to determine the change indicator characteristics in the detailed indicator characteristics.

[0045] Optionally, the above step S22 includes steps S221-S222, which are specifically: S221 uses the quantity proportion to determine the deviation of the quantity proportion in different intervals between adjacent unit time periods. When it is determined by the deviation that there is no deviation interval between different unit time periods and adjacent unit time periods, it is determined that the detailed indicator feature does not belong to the variable indicator feature. When there is a unit time period with a deviation interval between adjacent unit time periods, step S222 is entered; S222: obtaining the number of unit time periods with deviation intervals. When the number of unit time periods with deviation intervals is greater than the number of preset time periods, determining that the detailed indicator feature belongs to a variable indicator feature. When the number of unit time periods with deviation intervals is not greater than the number of preset time periods, entering step S223. S223 uses the average value of the deviation amount of the number proportion in different intervals and the number of deviation intervals to determine the characteristic variation coefficient of the adjacent unit time periods. When the number of unit time periods whose characteristic variation coefficient is greater than the preset variation coefficient threshold does not meet the requirements, it is determined that the detailed indicator feature belongs to the variation indicator feature. When the number of unit time periods whose characteristic variation coefficient is greater than the preset variation coefficient threshold meets the requirements, it proceeds to step S23.

[0046] Optionally, the above step S23 includes steps S231-S233, specifically: S231 determines the distribution change period in the unit period based on the characteristic change coefficients of different unit periods. When the number of distribution change periods does not meet the requirement, the detailed indicator feature is determined to be a change indicator feature. When the number of distribution change periods meets the requirement, the process proceeds to step S232. S232: When the number of distribution change periods is greater than the preset number of distribution change periods, proceed to step S223; when the number of distribution change periods is not greater than the preset number of distribution change periods, proceed to step S224; S233 obtains the average value of the characteristic variation coefficients of different distribution variation periods. When the average value of the characteristic variation coefficients is greater than the preset characteristic variation coefficient threshold, the detailed indicator feature is determined to be a variation indicator feature. When the average value of the characteristic variation coefficients is not greater than the preset characteristic variation coefficient threshold, the process proceeds to step S234. S234 determines a comprehensive variation coefficient based on the number of the distribution time periods and the characteristic variation coefficients of different distribution variation time periods, and uses the comprehensive variation coefficient to determine the variation indicator characteristics in the detailed indicator characteristics.

[0047] Furthermore, the method for determining the potential correlation index feature in the change index feature is: Determine the quantity proportion in different intervals within the unit time period based on the distribution data of the detailed indicator characteristics in different intervals within the unit time period; Determine the quantity proportion of the change index characteristics in different intervals in the unit time period within the abnormal time period of the index, and the deviation from other unit time periods to determine the interval with deviation from the quantity proportion in other unit time periods, and use it as the distribution deviation interval; Determine the deviation period ratio factor of different indicator abnormal periods by the proportion of the number of other unit periods in the distribution deviation interval; The characteristic distribution deviation coefficient of the changing indicator feature is determined by using the average value of the deviation period proportion factors of different indicator abnormal periods, and whether the changing indicator feature is a potential correlation indicator feature is determined according to the characteristic distribution deviation coefficient.

[0048] Specifically, when the feature distribution deviation coefficient is greater than a preset distribution deviation coefficient threshold, the change indicator feature is determined to be a potential correlation indicator feature.

[0049] It can be understood that when the deviation of the quantity proportion is greater than the preset proportion deviation threshold, the interval is determined to be a distribution deviation interval.

[0050] It should also be noted that the method for determining the potential correlation index feature in the change index feature includes steps S31-S34, specifically: S31 determines the quantity proportions in different intervals within the unit time period based on the distribution data of the detailed indicator characteristics in different intervals within the unit time period; S32 determines the quantity proportion of the change index feature in different intervals in the unit time period within the abnormal index period, and the deviation from other unit time periods determines the interval where the quantity proportion in other unit time periods deviates, and uses it as the distribution deviation interval; S33 determines the distribution deviation coefficient of the indicator abnormal period and other unit periods according to the number of the distribution deviation intervals and the deviation of the number ratio of different distribution deviation intervals, and determines the comprehensive deviation coefficient of different indicator abnormal periods according to the distribution deviation coefficients of different unit periods; S34 determines the characteristic distribution deviation coefficient of the changing indicator feature by using the average value of the comprehensive deviation coefficients of different indicator abnormal time periods, and determines whether the changing indicator feature is a potential correlation indicator feature according to the characteristic distribution deviation coefficient.

[0051] Optionally, the above step S32 includes steps S321-S322, which are specifically: S321 determines the quantity proportion of the change index feature in different intervals in the unit time period within the abnormal index period, and the deviation from other unit time periods determines the interval with deviation from the quantity proportion in other unit time periods, and uses it as the distribution deviation interval. When all abnormal index periods have distribution deviation intervals, enter step S322. When there is an abnormal index period without distribution deviation interval, it is determined that the change index feature does not belong to the potential correlation index feature; S322 When all the abnormal indicator periods have distribution deviation intervals with other unit periods, enter step S33. When there is an abnormal indicator period that does not have a distribution deviation interval with other unit periods, it is determined that the change indicator feature does not belong to the potential correlation indicator feature.

[0052] Optionally, the above step S33 includes steps S331-S333, specifically: S331 determines the distribution deviation coefficient between the abnormal indicator period and other unit periods by the number of the distribution deviation intervals and the deviation of the number of different distribution deviation intervals. When there is an abnormal indicator period whose distribution deviation coefficient with other unit periods is less than the preset coefficient threshold, enter step S332. When the distribution deviation coefficients of all abnormal indicator periods with other unit periods are not less than the preset coefficient threshold, it is determined that the change indicator feature belongs to the potential correlation indicator feature. S332 obtains the number of other unit time periods in which the distribution deviation coefficient is less than the preset coefficient threshold in different indicator abnormal time periods. When there are other unit time periods in which the distribution deviation coefficient is less than the preset coefficient threshold and the number of indicator abnormal time periods does not meet the requirement, it is determined that the change indicator feature does not belong to the potential correlation indicator feature. When there is no other unit time period in which the distribution deviation coefficient is less than the preset coefficient threshold and the number of indicator abnormal time periods does not meet the requirement, the process proceeds to step S333. S333 determines the comprehensive deviation coefficients of different indicator abnormal time periods based on the distribution deviation coefficients of different unit time periods. When there is an indicator abnormal time period with a comprehensive deviation coefficient less than the set deviation coefficient threshold, it is determined that the variable indicator feature does not belong to the potential associated indicator feature. When there is no indicator abnormal time period with a comprehensive deviation coefficient less than the set deviation coefficient threshold, proceed to step S34.

[0053] Specifically, Figure 4 As shown, the method for determining the associated indicator feature is: Based on the changes in the distribution data of the detailed indicator characteristics in different unit time periods with different characteristics of the potential correlation indicator and the changes in the credit indicators, the change correlation coefficients in different unit time periods are determined; Based on the change correlation coefficients in different unit time periods, the correlation coefficients of the potential correlation indicator features are determined, and the correlation indicator features in the potential correlation indicator features are determined based on the correlation coefficients.

[0054] Furthermore, the correlation indicator feature is a potential correlation indicator feature whose correlation coefficient is greater than a preset correlation coefficient threshold.

[0055] It should also be noted that the attribution analysis of changes in credit indicators specifically includes: Determining the order of the correlation indicator features based on the sorting results of the correlation coefficients of the correlation indicator features; Obtaining the abnormal time periods of the credit indicator within the specified time period, and determining the abnormal coefficient of the credit indicator by using the proportion of the number of the abnormal time periods within the specified time period; The number of attribution analyses of the credit indicator is determined according to the indicator abnormality coefficient of the credit indicator, and the change attribution analysis of the credit indicator is performed in the order of the number of attribution analyses and the characteristics of the associated indicators. Embodiment 2 In the second aspect, as Figure 5 As shown, the present invention provides a computer system, comprising: a memory and a processor connected in communication, and a computer program stored in the memory and capable of running on the processor, characterized in that: when the processor runs the computer program, it executes the above-mentioned attribution analysis method for changes in overdue indicators. Embodiment 3 In a third aspect, the present invention provides an attribution analysis model for overdue indicator changes, using the above-mentioned attribution analysis method for overdue indicator changes, specifically including: Focus on the target screening module, change feature screening module, correlation feature screening module, and attribution analysis processing module; The focus target screening module is responsible for determining the abnormal indicator period based on the change data of the credit indicator of the credit business within the specified period, and determining the focus credit indicator among the credit indicators in combination with the fluctuation of the credit characteristics between different unit time periods; The change feature screening module is responsible for dividing the credit application users of the credit business into multiple intervals based on the feature data under the detailed indicator feature, and determining the change indicator feature in the detailed indicator feature according to the change of the distribution data of the detailed indicator feature in different unit time periods in different intervals in the specified time period; The associated feature screening module is responsible for determining potential associated indicator features in the change indicator features by using the distribution data of different intervals of the change indicator features within the indicator abnormal period and the deviations within other unit time periods; The attribution analysis processing module is responsible for determining the correlation coefficients and correlation indicator characteristics of different potential correlation indicator characteristics based on the changes in the distribution data of the detailed indicator characteristics in different unit time periods of the potential correlation indicator characteristics in the specified time period and the changes in the credit indicators, and performing attribution analysis of the changes in credit indicators based on the sorting results of the correlation coefficients of the correlation indicator characteristics.

[0056] Each embodiment in this specification 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. In particular, for the device, equipment, and non-volatile computer storage medium embodiments, since they are basically similar to the method embodiments, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment.

[0057] The above is a description of a specific embodiment of the specification. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recorded in the claims can be performed in an order different from that in the embodiments and still achieve the desired results. In addition, the processes depicted in the drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0058] The above description is only one or more embodiments of this specification and is not intended to limit this specification. For those skilled in the art, one or more embodiments of this specification may have various changes and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of one or more embodiments of this specification shall be included in the scope of the claims of this specification.

Claims

1. A method for attribution analysis of changes in overdue indicators, characterized in that: Specifically include: The credit indicators of the credit business are used to determine the abnormal periods of indicators based on the change data within the specified period, and the credit indicators of concern are determined based on the fluctuation of credit characteristics between different unit periods; Based on the characteristic data under the detailed indicator characteristics, the credit application users of the credit business are divided into multiple intervals, and according to the change of the distribution data of the detailed indicator characteristics in different unit time periods in different intervals in the specified time period, the change indicator characteristics in the detailed indicator characteristics are determined; Determine potential related indicator features in the changing indicator features by using the deviation of distribution data of different intervals of the changing indicator features during the indicator abnormal period and other unit time periods; Based on the changes in the distribution data of the detailed indicator characteristics in different unit time periods of the potential correlation indicator characteristics in different intervals and the changes in credit indicators within the specified time period, the correlation coefficients and correlation indicator characteristics of different potential correlation indicator characteristics are determined, and the attribution analysis of the changes in credit indicators is performed based on the sorting results of the correlation coefficients of the correlation indicator characteristics.

2. The attribution analysis method for overdue index changes according to claim 1, characterized in that: The designated period is determined according to the number of credit application users of the credit service, wherein the greater the number of credit users of the credit service, the shorter the designated period is.

3. The attribution analysis method for overdue index changes according to claim 1, characterized in that: The credit indicators include collection rate, migration rate, recovery rate, non-performing loan rate and overdue rate.

4. The attribution analysis method for overdue index changes according to claim 1, characterized in that: The abnormal indicator period is a unit period when the credit indicator is not within the preset indicator range.

5. The attribution analysis method for overdue index changes according to claim 1, characterized in that: The method for determining the credit indicator of interest in the credit indicator is as follows: Obtaining the abnormal time periods of the credit indicator within the specified time period, and determining the abnormal coefficient of the credit indicator by using the proportion of the number of the abnormal time periods within the specified time period; Based on the fluctuation of credit characteristics between different unit time periods, the variation of credit characteristics between different adjacent unit time periods is determined, and the variation unit time period within the unit time period is determined according to the variation, and the index variation coefficient of the credit index is determined according to the proportion of the variation time period within the specified time period; The attention coefficient of the credit indicator is determined based on the indicator variation coefficient and the indicator abnormality coefficient, and the attention credit indicator in the credit indicator is determined through the attention coefficient.

6. The attribution analysis method for overdue index changes according to claim 5, characterized in that: The credit indicator of concern is a credit indicator whose concern coefficient is greater than a preset coefficient threshold.

7. The attribution analysis method for overdue index changes according to claim 1, characterized in that: The credit application users of the credit business are divided into multiple sections, including: According to the interval in which the characteristic data of the credit applicant is located under the detailed indicator characteristics, the credit applicant is divided into a corresponding interval.

8. The attribution analysis method for overdue index changes according to claim 1, characterized in that: The method for determining the change indicator feature in the detailed indicator feature is: Determine the quantity proportion in different intervals within the unit time period based on the distribution data of the detailed indicator characteristics in different intervals within the unit time period; Determine the deviation interval between adjacent unit time periods by using the quantity proportion, and determine the characteristic variation coefficient between adjacent unit time periods according to the quantity of the deviation interval; The distribution change period in the unit period is determined based on the characteristic change coefficient, and the change index feature in the detailed index feature is determined using the number of the distribution change periods.

9. A computer system comprising: A memory and a processor that are communicatively connected, and a computer program stored in the memory and capable of running on the processor, characterized in that: when the processor runs the computer program, it executes an attribution analysis method for changes in overdue indicators as described in any one of claims 1-8.

Images