CN112184415A - Data processing method and device, electronic equipment and storage medium - Google Patents
Data processing method and device, electronic equipment and storage medium Download PDFInfo
- Publication number
- CN112184415A CN112184415A CN202011019016.9A CN202011019016A CN112184415A CN 112184415 A CN112184415 A CN 112184415A CN 202011019016 A CN202011019016 A CN 202011019016A CN 112184415 A CN112184415 A CN 112184415A
- Authority
- CN
- China
- Prior art keywords
- historical
- macroscopic
- exposure
- target
- data
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q40/00—Finance; Insurance; Tax strategies; Processing of corporate or income taxes
- G06Q40/03—Credit; Loans; Processing thereof
Landscapes
- Business, Economics & Management (AREA)
- Accounting & Taxation (AREA)
- Finance (AREA)
- Engineering & Computer Science (AREA)
- Development Economics (AREA)
- Economics (AREA)
- Marketing (AREA)
- Strategic Management (AREA)
- Technology Law (AREA)
- Physics & Mathematics (AREA)
- General Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The embodiment of the invention discloses a data processing method, a data processing device, electronic equipment and a storage medium, wherein the method comprises the following steps: acquiring a macroscopic year historical value, a historical numerical value of a target macroscopic factor and historical data of exposure; constructing a macroscopic prediction model according to the historical value of the macroscopic year and the historical numerical value of the target macroscopic factor; verifying the risk parameters of the exposure according to historical data of the exposure based on a macroscopic prediction model; by the technical scheme, the purpose of automatically verifying the macroscopic prediction model is achieved.
Description
Technical Field
The embodiment of the invention relates to the technical field of computers, in particular to a data processing method and device, electronic equipment and a storage medium.
Background
Model risk management is a hot problem in the field of risk management of financial institutions at home and abroad. Existing model risk management organizations may conduct the work associated with developing model validation.
Currently, the business department usually adopts manual verification. According to the requirements of new financial tool criteria, all financial assets in a subtraction range need to be subjected to value-by-value increasing and decreasing preparation, the difficulty of manually adjusting model risk parameters is high, the execution is very complex, and time and labor are wasted. In addition, the model verification process is accompanied by huge workload, errors are easy to generate, and improvement is urgently needed.
Disclosure of Invention
The invention provides a data processing method, a data processing device, electronic equipment and a storage medium, which aim to automatically verify a macro prediction model.
In a first aspect, an embodiment of the present invention provides a data processing method, where the method includes:
acquiring a macroscopic year historical value, a historical numerical value of a target macroscopic factor and historical data of exposure;
constructing a macroscopic prediction model according to the historical value of the macroscopic year and the historical numerical value of the target macroscopic factor;
and verifying the exposure risk parameters according to the exposure historical data based on a macroscopic prediction model.
In a second aspect, an embodiment of the present invention further provides a data processing apparatus, where the apparatus includes:
the data acquisition module is used for acquiring the historical value of the macroscopic year, the historical numerical value of the target macroscopic factor and the historical data of the exposure;
the model construction module is used for constructing a macroscopic prediction model according to the macroscopic year historical value and the historical numerical value of the target macroscopic factor;
and the verification result output module is used for verifying the exposure risk parameters according to the exposure historical data based on the macroscopic prediction model.
In a third aspect, an embodiment of the present invention further provides an electronic device, where the electronic device includes:
one or more processors;
a storage device for storing one or more programs,
when the one or more programs are executed by the one or more processors, the one or more processors implement the data processing method according to any one of the embodiments of the present invention.
In a fourth aspect, the embodiment of the present invention further provides a computer-readable storage medium, on which a computer program is stored, and the computer program, when executed by a processor, implements the data processing method according to any one of the embodiments of the present invention.
According to the technical scheme of the embodiment, a macroscopic prediction model is constructed according to the acquired macroscopic year historical value and the historical numerical value of the target macroscopic factor; and further, according to the historical data of the exposure, verifying and calculating the risk parameters of the exposure by using a macroscopic prediction model so as to verify the accuracy of the prediction result of the macroscopic prediction model. Compared with the existing macro prediction model verification scheme, the embodiment greatly saves human resources, improves the model prediction accuracy, and provides a new idea for automatically verifying the macro prediction model.
Drawings
Fig. 1 is a flowchart of a data processing method according to an embodiment of the present invention;
fig. 2 is a flowchart of a data processing method according to a second embodiment of the present invention;
fig. 3 is a flowchart of a data processing method according to a third embodiment of the present invention;
fig. 4 is a flowchart of a data processing method according to a fourth embodiment of the present invention;
fig. 5 is a schematic structural diagram of a data processing apparatus according to a fifth embodiment of the present invention;
fig. 6 is a schematic structural diagram of an electronic device according to a sixth embodiment of the present invention.
Detailed Description
The present invention will be described in further detail with reference to the accompanying drawings and examples. It is to be understood that the specific embodiments described herein are merely illustrative of the invention and are not limiting of the invention. It should be further noted that, for the convenience of description, only some of the structures related to the present invention are shown in the drawings, not all of the structures.
Example one
Fig. 1 is a flowchart of a data processing method according to an embodiment of the present invention. The embodiment can be applied to the condition of verifying the constructed macro prediction model. The method may be implemented by a data processing apparatus, which may be implemented by means of software and/or hardware. As shown in fig. 1, the data processing method provided in the embodiment of the present invention specifically includes:
s110, acquiring historical values of the macroscopic year, historical numerical values of the target macroscopic factors and historical data of exposure.
Wherein, the macroscopic year historical value refers to the historical numerical value of the macroscopic year value, and the macroscopic year value refers to the coefficient value which comprehensively reflects a plurality of external macroscopic factors.
The target macroscopic factor is selected from a plurality of macroscopic factors and is required for constructing a macroscopic prediction model. Common macroscopic factors can comprise total domestic production value, consumption price index, unemployment rate, annual interest rate, interest borrowing rate and the like, and can also comprise derivative variable factors generated by leading and lagging the macroscopic factors. Optionally, the embodiment may randomly extract a preset number of macro factors from the multiple macro factors as the target macro factor; further, in order to ensure the accuracy of the subsequent model prediction, the target macroscopic factor can also be determined by the following way: acquiring a reference prediction model provided by a user from a visual interface; extracting at least two candidate macroscopic factors from the reference prediction model; and selecting a target macroscopic factor from the at least two candidate macroscopic factors according to the historical distribution condition of the debt items in the exposure and the historical values of the at least two candidate macroscopic factors.
The reference prediction model is a model which is selected to have high score through carrying out statistical analysis and scoring on a macroscopic prediction model with a plurality of possible macroscopic factors, and manually screening, and is used as the reference prediction model. And calculating a correlation coefficient of credit risk exposure according to the historical distribution condition of the open debt items, and selecting a target macroscopic factor from the macroscopic factors of the reference prediction model.
It can be understood that the target macroscopic factor is selected from the macroscopic factors of the reference prediction model, the selection range of the macroscopic factor is narrowed, and the target macroscopic factor is selected according to the correlation coefficient of the credit risk exposure, so that the effect of further improving the data processing accuracy is achieved.
The open is a pool-divided open with credit risk exposure, and each financial institution can divide the financial assets according to the demands of the financial institutions on the fineness of the divided financial assets and the quantity condition to obtain each pool-divided open. The pool-divided opening comprises a plurality of debts (further loan debts), and the debts in the opening can be divided into a plurality of credit levels according to a five-level loan classification system. The historical data of the exposure refers to the historical time point detail data of the debt in the exposure, and can include but is not limited to the historical default rate of the debt, the credit rating of the debt and the like.
Optionally, the embodiment may provide a visual interface, and then the historical macro-year values and the historical values of the target macro-factors may be obtained by importing data through a front-end interface of the system; the historical data of the exposure is obtained by extracting from a database of the correlation system. And after the data are acquired, storing the basic data for verifying the macroscopic prediction model into a database.
And S120, constructing a macroscopic prediction model according to the macroscopic year historical value and the historical numerical value of the target macroscopic factor.
Wherein, the macroscopic prediction model is used for expressing the relation between the macroscopic annual number value and the target macroscopic factor.
In this embodiment, a macro prediction model may be constructed by performing multiple regression (e.g., ternary regression) analysis on the macro age value and the target macro factor and using a least square method.
Optionally, after the macro prediction model is built, the macro performance of the built macro prediction model may be directly verified, which specifically may be: and evaluating the macro prediction model according to the target evaluation index.
In this embodiment, the target evaluation index may be selected from an index library according to actual needs, or may be default by the system; the target evaluation index comprises at least one of a significance test index, a goodness-of-fit test index and a residual error test index.
Specifically, according to the target evaluation index, evaluating the macro prediction model may include verifying parameters in the constructed macro prediction model, for example, performing significance check on a target macro factor; verification of the fitting result of the macro prediction model, such as residual error test and goodness-of-fit test, can also be included.
And S130, verifying the exposure risk parameters according to the exposure historical data based on the macroscopic prediction model.
In the embodiment, considering that the exposure risk parameter is influenced by the output value of the macroscopic prediction model, the purpose of verifying the accuracy of the macroscopic prediction model prediction is indirectly realized by verifying the exposure risk parameter. Wherein, the risk parameters of exposure include, but are not limited to, default rate, default loss rate, default risk exposure, etc.
In this embodiment, based on the macroscopic prediction model, according to the exposure historical data, the verification of the exposure risk parameter may specifically be: estimating the open risk parameter according to the macroscopic annual value obtained by the macroscopic prediction model, comparing the estimated risk parameter with the actual risk parameter, and determining the deviation degree between the estimated risk parameter and the actual risk parameter; and according to the deviation result, evaluating the accuracy of risk parameter estimation, and further realizing the verification of the open risk parameter.
In order to further improve the user experience, for example, the verification result may be visually displayed through a system front-end interface according to the calculation result of the model verification. After the verification process is completed, performing visual display based on the result generated by verification, such as displaying with tools such as a table, a broken line graph and a pie graph, and accordingly providing download of all verification results, intermediate data and basic data, such as providing download of relevant verification reports of word and Excel.
Preferably, the system architecture of the macro-prediction model verification in this embodiment is programmed and implemented by using an open source product of PYTHON, and accordingly, a flexible parameter configuration management function is provided.
According to the technical scheme of the embodiment, a macroscopic prediction model is constructed according to the acquired macroscopic year historical value and the historical numerical value of the target macroscopic factor; and further, according to the historical data of the exposure, verifying and calculating the risk parameters of the exposure by using a macroscopic prediction model so as to verify the accuracy of the prediction result of the macroscopic prediction model. Compared with the existing macro prediction model verification scheme, the embodiment greatly saves human resources, improves the model prediction accuracy, and provides a new idea for automatically verifying the macro prediction model.
Example two
Fig. 2 is a flowchart of a data processing method according to a second embodiment of the present invention. The present embodiment is further described in detail with respect to a process of verifying exposure risk parameters by applying a macro prediction model based on the above embodiments. Referring to fig. 2, a data processing method provided in this embodiment includes:
s210, acquiring historical values of the macroscopic year, historical numerical values of the target macroscopic factors and historical data of exposure.
And S220, constructing a macroscopic prediction model according to the macroscopic year historical value and the historical numerical value of the target macroscopic factor.
And S230, determining the predicted value of the target macroscopic factor according to the historical numerical value of the target macroscopic factor.
Wherein, the predicted value of the target macroscopic factor refers to the future value of the target macroscopic factor.
Optionally, according to the historical numerical value of the target macroscopic factor, performing time series analysis to obtain a prediction model of the target macroscopic factor; and predicting the target macroscopic factor according to the prediction model of the target macroscopic factor to obtain the predicted value of the target macroscopic factor.
And S240, determining a macroscopic year predicted value according to the macroscopic prediction model and the predicted value of the target macroscopic factor.
Wherein, the macroscopic year prediction value refers to the prediction value of the macroscopic year value. In this embodiment, the predicted value of the target macroscopic factor is input into the macroscopic prediction model, so as to obtain the macroscopic year predicted value.
And S250, obtaining the prospective default rate of the opening according to the macroscopic year predicted value and the historical default rate of the opening based on the Nootn model.
The prospective default rate of the exposure is the default rate of the exposure at a certain future time node. The historical default rate of exposure refers to the historical default rate of exposure at each time node in the past.
The norton model is a tool for quantifying credit risks, and the specific norton model comprises the relationship among a long-term default rate, a short-term default rate and a macroscopic annual value.
Specifically, the long-term default rate of the exposure can be determined by performing statistical analysis on the historical default rate of each time node in the past of the exposure; and inputting the macroscopic year predicted value and the open long-term default rate into a Norton model, and further obtaining the open future predicted short-term default rate, namely the open forward-looking default rate in the embodiment.
And S260, carrying out statistical test according to the historical default rate and the look-ahead default rate of the exposure.
Wherein, the statistical test is to test the deviation degree between the future default rate and the historical default rate based on a statistical test method.
Alternatively, the statistical test may be specifically a traffic light binomial distribution test method. Assuming that the deviation numbers of the historical default rate obey binomial distribution, the deviation numbers which can be accepted under the confidence intervals of 95 percent and 99.99 percent, namely the yellow light triggering critical value and the red light triggering critical value, are determined based on the total number of the predicted points of the prospective default rate and the 10 percent deviation probability. If the number of the deviation of the statistical test is smaller than the critical value of the yellow light, the accuracy of predicting the open prospective default rate is high.
According to the technical scheme of the embodiment, the prospective default rate of the opening can be automatically and accurately predicted based on the macroscopic annual value predicted by the macroscopic prediction model and the historical default rate of the opening by combining the Noton model; comparing the predicted prospective default rate of the opening with the historical default rate, and verifying the accuracy of the risk parameter estimation by adopting a statistical test method; through the technical scheme, the embodiment provides an optional mode for verifying the open risk parameters by applying the macroscopic prediction model.
Optionally, based on the norton model, after obtaining the prospective default rate of the exposure according to the macroscopic year prediction value and the historical default rate of the exposure, the method further includes:
A. determining the average default rate of the exposure according to the historical default rate of the debt items in the historical data of the exposure;
in this embodiment, historical default rates of a plurality of debt items included in the exposure may be averaged, and then the average default rate of the exposure is obtained.
As an optional manner of this embodiment, according to the historical default rate of the debt items in the historical data of the exposure, determining the average default rate of the exposure may be: selecting target debt from at least two debts included in the exposure according to the credit rating of the at least two debts in the historical data of the exposure; determining an upper limit value and a lower limit value of the historical default rate of the target debt item according to the historical default rate of the target debt item; and determining the average default rate of the exposure according to the upper limit value and the lower limit value of the historical default rate of the target debt items.
The target debt items refer to a preset number of debt items selected from a plurality of (i.e., two or more) debt items included in the exposure according to the credit level distribution condition of the debt items in the exposure, for example, the debt items may be sorted in a descending order according to the credit level, and then the debt items sorted in the previous preset number may be selected as the target debt items according to the sorting result. And counting the obtained historical default rate of the target debt item, and determining the maximum historical default rate and the minimum historical default rate, namely the upper limit value and the lower limit value of the corresponding historical default rate. And carrying out extremum processing on the historical default rate of all debt items under the exposure according to the upper limit value and the lower limit value of the historical default rate. After all debt items to be uncovered are subjected to extreme value processing, a weighted average mode can be selected to determine the average default rate of the uncovering.
Optionally, in the extremum processing in this embodiment, the historical default rate of each debt item in the exposure may be determined according to the upper limit value and the lower limit value of the historical default rate; and adjusting the historical default rate of the debt items with the historical default rate larger than the upper limit value of the historical default rate to be the upper limit value of the historical default rate, and adjusting the historical default rate of the debt items with the historical default rate smaller than the upper limit value of the historical default rate in the exposure to be the lower limit value of the historical default rate.
B. Determining a macroscopic annual evaluation value according to the open look-ahead default rate and the average default rate on the basis of a Norton model;
the macroscopic annual internal evaluation value refers to a macroscopic annual value calculated under an internal evaluation method. The internal evaluation method is a method for calculating the macroscopic annual internal evaluation value by inputting the calculated average default rate and the uncovered prospective default rate into a Norton model.
C. And updating the historical default rate of the uncovered debt items according to the macroscopic annual internal evaluation value based on the Nonton model.
The updating of the historical default rate of the open debt items means that the historical default rate of the open debt items is adjusted to obtain the prospective default rate of the open debt items.
The technical scheme of the embodiment is based on the Nonton model, the historical default rate of the open debt items is updated according to the macroscopic annual evaluation value, the adjusted prospective default rate of the open debt items is obtained, and the effect of providing basic data support and follow-up research and application for the credit risk management of the financial institution is achieved by performing measure calculation on the prospective default rate of the open debt items.
EXAMPLE III
Fig. 3 is a flowchart of a data processing method according to a third embodiment of the present invention. The embodiment is added with a process of detecting the distribution of the debt items in the exposure on the basis of the embodiment. Referring to fig. 3, a data processing method provided in this embodiment includes:
s310, acquiring historical values of the macroscopic year, historical numerical values of the target macroscopic factors and historical data of exposure.
And S320, constructing a macroscopic prediction model according to the macroscopic year historical value and the historical numerical value of the target macroscopic factor.
S330, verifying the exposure risk parameters according to the exposure historical data based on the macroscopic prediction model.
And S340, determining a comparison time point according to the time point to be measured.
Wherein, the time point to be measured is an initial time point determined by carrying out the distribution test of the debt items in the exposure.
The comparison time point is a historical time point determined before the measured time point by a certain time. Wherein, the period of time can be determined according to the requirement of monitoring the debt in the exposure, such as half a year or a year.
S350, determining first to-be-tested data and second to-be-tested data related to the target exposure according to the to-be-tested time point, the comparison time point, the target exposure identification and the exposure historical data.
The target exposure id is an identifier for uniquely identifying a specific exposure, and may be, for example, a number of the target exposure. The first data to be tested is related data of the debt item of the target exposure acquired from the historical data of the exposure according to the currently provided time point to be tested, and specifically may include the number of the debt item and the default number of the debt item included at the time point to be tested. Correspondingly, the second data to be measured refers to the related data of the debt items of the same target exposure acquired from the historical data of the exposure according to the currently provided comparison time point, and specifically may include the number of the debt items and the default number of the debt items included at the comparison time point.
Furthermore, related data of the debt items included in the designated credit level under the target exposure can be acquired, and the first data to be tested can be related data of the debt items belonging to a certain credit level under the target exposure acquired from historical data of the exposure according to the currently provided time point to be tested; correspondingly, the second data to be measured is related data of debt items belonging to a certain credit grade under the same target exposure acquired from the historical data of the exposure according to the currently provided comparison time point.
Specifically, historical data of the target exposure can be obtained from the historical data of the exposure according to the target exposure identifier; acquiring first to-be-detected data related to the target exposure according to the to-be-detected time point and the historical data of the target exposure; similarly, second data to be measured related to the target exposure can be obtained according to the comparison time point and the historical data of the target exposure.
And S360, carrying out stability index inspection according to the first data to be tested and the second data to be tested.
The stability index inspection is to calculate the stability index according to the number of debt items exposed (belonging to a certain credit grade in further exposure) at two time points (a time point to be measured and a comparison time point) before and after and the change condition of the default number of the debt items; utilize the stability index to examine debt at open distribution condition, when open stability index is unusual, show open mobility increase, need consider the open rationality in minute pond this moment to further adjust open structure in view of the above. A
According to the technical scheme of the embodiment, according to the history data of the opening, after the time for measuring the opening is determined, the stability index inspection of the opening is carried out, the actual default deviation condition of the debt item in the opening is analyzed through the stability index inspection, the purpose of inspecting and monitoring the stability of the opening debt item is achieved, and theoretical support is further provided for a financial institution to adjust the opening structure and issue more excellent risk countermeasures.
Optionally, after obtaining the historical values of the macroscopic year, the historical values of the target macroscopic factor and the historical data of the exposure, the exposure can be subjected to a kolmogorov index test. The method specifically comprises the following steps: determining debt default quantity and debt non-default quantity related to the credit rating of the target exposure according to the historical data of the exposure, the target exposure identifier and the duration to be measured; and carrying out Kolmogorov index test according to the debt default quantity and the debt non-default quantity which are related to the credit rating in the target exposure.
The duration to be measured refers to a period of time between the time point to be measured and the time point to be compared. The Kolmogorov index test means that in a fixed time length to be tested, the number of default and number of non-default under the open credit level are counted, and the Kolmogorov index value is calculated according to the number of default and number of non-default; and evaluating the risk ranking value of the credit rating of the financial institution by using the calculated Kolmogorov index to check whether the credit rating is reasonable from high to low corresponding to the actual default value.
Optionally, after the macroscopic year historical value, the historical numerical value of the target macroscopic factor and the historical data of the exposure are acquired, the chiny index inspection and the information value index inspection can be performed according to the debt default quantity and the debt non-default quantity associated with the credit rating of the target exposure.
According to the technical scheme of the embodiment, according to the historical data of the exposure, after the time for measuring the exposure is determined, the Kolmogorov index inspection of the exposure is carried out, and through the Kolmogorov index inspection, whether the default value corresponding to the credit rating from high to low is reasonable or not is analyzed, so that the purpose of inspecting and monitoring the risk sequencing capacity of the exposure debt items is achieved, and theoretical support is further provided for a financial institution to adjust the exposure structure and provide more excellent risk countermeasures.
Optionally, after acquiring the historical macroscopic year values, the historical values of the target macroscopic factors and the historical data of the exposure, other index tests can be performed on the exposure, including but not limited to performing debt concentration index test, fluidity index test, homogeneity and heterogeneity index test on the exposure.
Example four
Fig. 4 is a schematic diagram of a data processing procedure according to a fourth embodiment of the present invention. The present embodiment is a preferred example provided on the basis of the above-described embodiments. The example also involves verifying the macroscopic prediction model, the exposure risk parameter, and the exposure debt distribution. With reference to fig. 4, a data processing method provided in this embodiment includes: the system comprises a data index maintenance and data processing part, a risk parameter verification part, a macroscopic prediction model verification part, an open debt item distribution inspection part and a verification result output part.
Wherein, the data index maintenance and data processing part is used for: maintaining a macroscopic year historical value, a macroscopic factor index and a reference prediction model parameter; according to the business consultation scheme, processing default rate of the historical time sequence of basic debt items by a GP database, and then processing detailed data of the historical time of the uncovered debt items from the GP database; and pushing the processed and maintained data into a corresponding ORACLE database.
The risk parameter verification section is configured to: calculating a correlation coefficient of credit risk exposure according to the historical time sequence and the exposure classification of the exposure debt items; extracting reference prediction model parameters, a macroscopic factor table and a macroscopic factor historical value; carrying out ternary regression on model parameters related to the opening and the historical time points to obtain a macroscopic prediction model, and further predicting a macroscopic annual value; calculating the prospective default rate of the corresponding opening by adopting the macroscopic annual value predicted by the Nonton model and the macroscopic prediction model; and (4) carrying out binomial traffic light inspection according to the exposure and the historical time point of the debt.
Further, the default rate of the uncovered debt items is adjusted by utilizing the uncovered prospective default rate obtained by the risk parameter verification part. The specific adjustment process may be: carrying out extremum processing of default rate on the debt items outside the historical open 90% data distribution credit level; after the opening is subjected to grading adjustment, calculating the average default rate of the graded opening according to the debt items in the opening in a weighted average mode; the average default rate of the classified opening and the prospective default rate of the opening are brought into a Norton model to obtain a macroscopic annual internal evaluation value based on an internal evaluation method; and inputting the macroscopic annual internal evaluation value based on the internal evaluation method into a Noton model, adjusting the default rate of each debt item under the opening, and finally obtaining the default rate of each debt item after prospective adjustment.
The macro-prediction model verification part is used for: performing parameter inspection on the macro prediction model obtained by the multiple regression calculation, such as macroscopic factor significance inspection, sensitivity test, macro prediction model independent variable collinearity test and the like; carrying out residual error detection on the macro prediction model, such as stability test of model residual error, autocorrelation test of model residual error, heteroscedasticity test of model residual error, normality prediction of model residual error and the like; and (5) carrying out goodness-of-fit inspection on the macroscopic prediction model.
The open debt distribution inspection part is used for: comparing the migration change of 12 months before the debt according to the provided current historical time point, and performing stability index test, flow index coefficient test and pool-dividing risk sequencing capability test (including Gini index test, information value test, Kolmogorov test and the like); according to the distribution of the open debt items, performing a pool-dividing debt item concentration degree test; and performing homogeneity and heterogeneity comparison and verification according to the open pool dividing condition.
The verification result output part is used for: after the verification process is completed, performing visual display based on the result generated by verification, such as displaying with tools such as a table, a broken line graph and a pie graph, and accordingly providing download of all verification results, intermediate data and basic data, such as providing download of relevant verification reports of word and Excel.
In the technical scheme of the embodiment, based on the method for verifying the macroscopic prediction model realized on line, the following three aspects of verification are performed: constructing a macroscopic prediction model according to basic data obtained by the data index maintenance and data processing part, and inspecting the macroscopic expression of the macroscopic prediction model; calculating the open prospective default rate according to the macroscopic annual value predicted by the macroscopic prediction model, carrying out binomial traffic light inspection, and simultaneously adjusting the prospective default rate of the historical debt items according to the predicted macroscopic annual value; according to the historical data of the exposure, the distribution condition of the debt items in the exposure is checked; by the technical scheme, human resources are greatly saved, and the accuracy of the macroscopic prediction model prediction is improved by carrying out relevant verification work on the model.
EXAMPLE five
Fig. 5 is a schematic structural diagram of a data processing apparatus according to an embodiment of the present invention. Referring to fig. 5, a data processing apparatus provided in an embodiment of the present application includes: a data acquisition module 501, a model construction module 502 and a verification result output module 503.
The data acquisition module 501 is configured to acquire a macroscopic year historical value, a historical numerical value of a target macroscopic factor, and historical data of an exposure;
a model construction module 502, configured to construct a macro prediction model according to the historical macro year values and the historical values of the target macro factors;
and the verification result output module 503 is configured to verify the exposure risk parameter according to the historical data of the exposure based on the macroscopic prediction model.
According to the technical scheme of the embodiment, a macroscopic prediction model is constructed according to the acquired macroscopic year historical value and the historical numerical value of the target macroscopic factor; and further, according to the historical data of the exposure, verifying and calculating the risk parameters of the exposure by using a macroscopic prediction model so as to verify the accuracy of the prediction result of the macroscopic prediction model. Compared with the existing macro prediction model verification scheme, the embodiment greatly saves human resources, improves the model prediction accuracy, and provides a new idea for automatically verifying the macro prediction model.
Further, the above apparatus further comprises:
the reference model acquisition module is used for acquiring a reference prediction model provided by a user from a visual interface before acquiring the historical values of the macroscopic year, the historical numerical values of the target macroscopic factors and the historical data of the exposure;
a candidate factor extraction module for extracting at least two candidate macroscopic factors from the reference prediction model;
and the target factor extraction module is used for selecting the target macroscopic factor from the at least two candidate macroscopic factors according to the historical distribution condition of the exposure debt and the historical numerical values of the at least two candidate macroscopic factors.
Further, the verification result output module comprises:
the factor predicted value confirmation unit is used for determining the predicted value of the target macroscopic factor according to the historical numerical value of the target macroscopic factor;
and the statistical test unit is used for carrying out statistical test on the macroscopic prediction model according to the predicted value of the target macroscopic factor and the historical default rate of the exposure in the historical data of the exposure.
Further, the statistical test unit is specifically configured to:
determining a macroscopic year predicted value according to the macroscopic prediction model and the predicted value of the target macroscopic factor;
based on a Nootn model, obtaining an open prospective default rate according to a macroscopic year predicted value and an open historical default rate;
and carrying out statistical test according to the historical default rate and the prospective default rate of the exposure.
Further, the verification result output module further includes:
the mean value determining unit is used for determining the average default rate of the exposure according to the historical default rate of the exposure debt items in the historical data of the exposure after obtaining the prospective default rate of the exposure according to the macroscopic year predicted value and the historical default rate of the exposure based on the Nonton model;
the internal evaluation value determining unit is used for determining a macroscopic annual internal evaluation value according to the open look-ahead default rate and the average default rate on the basis of the Noton model;
and the default rate updating unit is used for updating the historical default rate of the open debt items according to the macroscopic annual evaluation value based on the Nonton model.
Further, the mean value determining unit includes:
the debt item selecting subunit is used for selecting a target debt item from at least two debt items included in the exposure according to the credit levels of the at least two debt items in the historical data of the exposure;
the limit value determining subunit is used for determining the upper limit value and the lower limit value of the historical default rate of the target debt item according to the historical default rate of the target debt item;
and the mean value determining subunit is used for determining the average default rate of the exposure according to the upper limit value and the lower limit value of the historical default rate of the target debt items.
Further, the mean determination subunit is specifically configured to:
and carrying out extreme value processing on the debt items in the exposure according to the upper limit value and the lower limit value of the historical default rate of the target debt items to obtain the average default rate of the exposure.
Further, the above apparatus further comprises:
the comparison time point determining module is used for determining a comparison time point according to the time point to be measured after acquiring the historical value of the macroscopic year, the historical numerical value of the target macroscopic factor and the historical data of the exposure;
the to-be-detected data determining module is used for determining first to-be-detected data and second to-be-detected data related to the target opening according to the to-be-detected time point, the comparison time point, the target opening identification and the history data of the opening; the first data to be tested comprises the debt item quantity and the debt default quantity contained at the time point to be tested, and the second data to be tested comprises the debt item quantity and the debt default quantity contained at the comparison time point;
and the stability inspection module is used for inspecting the stability index according to the first data to be tested and the second data to be tested.
Further, the above apparatus further comprises:
the basic data acquisition module is used for determining debt default quantity and debt non-default quantity related to the credit rating of the target exposure according to the history data of the exposure, the target exposure identification and the duration to be measured after acquiring the history value of the macroscopic year, the history numerical value of the target macroscopic factor and the history data of the exposure;
and the risk inspection module is used for carrying out Kolmogorov index inspection according to debt default quantity and debt non-default quantity related to the credit rating of the target exposure.
Further, the above apparatus further comprises:
the prediction model verification module is used for evaluating the macro prediction model according to the target evaluation index after the macro prediction model is constructed according to the macro year historical value and the historical numerical value of the target macro factor; wherein the target evaluation index comprises at least one of a significance test index and a goodness-of-fit test index.
The data processing device provided by the embodiment of the invention can execute the data processing method provided by any embodiment of the invention, and has corresponding functional modules and beneficial effects of the execution method.
EXAMPLE six
Fig. 6 is a schematic structural diagram of an electronic device according to a sixth embodiment of the present invention, as shown in fig. 6, the electronic device includes a processor 60, a memory 61, an input device 62, and an output device 63; the number of processors 60 in the device may be one or more, and one processor 60 is taken as an example in fig. 6; the processor 60, the memory 61, the input device 62 and the output device 63 in the apparatus may be connected by a bus or other means, as exemplified by the bus connection in fig. 6.
The memory 61, which is a computer-readable storage medium, may be used to store software programs, computer-executable programs, and modules, such as program instructions/modules corresponding to the data processing method in the embodiment of the present invention (for example, the data acquisition module 501, the model construction module 502, and the verification result output module 503 in the data processing apparatus). The processor 60 executes various functional applications of the device and data processing, i.e., implements the above-described data processing method, by executing software programs, instructions, and modules stored in the memory 61.
The memory 61 may mainly include a program storage area and a data storage area, wherein the program storage area may store an operating system, an application program required for at least one function; the storage data area may store data created according to the use of the terminal, and the like. Further, the memory 61 may include high speed random access memory, and may also include non-volatile memory, such as at least one magnetic disk storage device, flash memory device, or other non-volatile solid state storage device. In some examples, the memory 61 may further include memory located remotely from the processor 60, which may be connected to the device over a network. Examples of such networks include, but are not limited to, the internet, intranets, local area networks, mobile communication networks, and combinations thereof.
The input device 62 may be used to receive entered numeric or character information and to generate key signal inputs relating to user settings and function controls of the apparatus. The output device 63 may include a display device such as a display screen.
EXAMPLE seven
An embodiment of the present invention further provides a storage medium containing computer-executable instructions, which when executed by a computer processor, are configured to perform a data processing method, including:
acquiring a macroscopic year historical value, a historical numerical value of a target macroscopic factor and historical data of exposure;
constructing a macroscopic prediction model according to the historical value of the macroscopic year and the historical numerical value of the target macroscopic factor;
and verifying the exposure risk parameters according to the exposure historical data based on a macroscopic prediction model.
Of course, the storage medium provided by the embodiment of the present invention contains computer-executable instructions, and the computer-executable instructions are not limited to the above method operations, and may also perform related operations in the data processing method provided by any embodiment of the present invention.
From the above description of the embodiments, it is obvious for those skilled in the art that the present invention can be implemented by software and necessary general hardware, and certainly, can also be implemented by hardware, but the former is a better embodiment in many cases. Based on such understanding, the technical solutions of the present invention may be embodied in the form of a software product, which can be stored in a computer-readable storage medium, such as a floppy disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a FLASH Memory (FLASH), a hard disk or an optical disk of a computer, and includes several instructions for enabling a computer device (which may be a personal computer, a server, or a network device) to execute the methods according to the embodiments of the present invention.
It should be noted that, in the embodiment of the data processing apparatus, the included units and modules are merely divided according to functional logic, but are not limited to the above division as long as the corresponding functions can be implemented; in addition, specific names of the functional units are only for convenience of distinguishing from each other, and are not used for limiting the protection scope of the present invention.
It is to be noted that the foregoing is only illustrative of the preferred embodiments of the present invention and the technical principles employed. It will be understood by those skilled in the art that the present invention is not limited to the particular embodiments described herein, but is capable of various obvious changes, rearrangements and substitutions as will now become apparent to those skilled in the art without departing from the scope of the invention. Therefore, although the present invention has been described in greater detail by the above embodiments, the present invention is not limited to the above embodiments, and may include other equivalent embodiments without departing from the spirit of the present invention, and the scope of the present invention is determined by the scope of the appended claims.
Claims (15)
1. A data processing method, comprising:
acquiring a macroscopic year historical value, a historical numerical value of a target macroscopic factor and historical data of exposure;
constructing a macroscopic prediction model according to the macroscopic year historical value and the historical numerical value of the target macroscopic factor;
and verifying the risk parameters of the exposure according to the historical data of the exposure based on the macroscopic prediction model.
2. The method of claim 1, wherein before obtaining the historical values of the macroscopic year, the historical values of the target macroscopic factors, and the historical data of the exposure, the method further comprises:
acquiring a reference prediction model provided by a user from a visual interface;
extracting at least two candidate macroscopic factors from the reference prediction model;
and selecting a target macroscopic factor from the at least two candidate macroscopic factors according to the historical distribution condition of the debt items in the exposure and the historical numerical values of the at least two candidate macroscopic factors.
3. The method of claim 1, wherein verifying risk parameters of an exposure based on the macro-predictive model from historical data of the exposure comprises:
determining a predicted value of the target macroscopic factor according to the historical numerical value of the target macroscopic factor;
and based on the macroscopic prediction model, carrying out statistical inspection on the exposure risk parameters according to the predicted value of the target macroscopic factor and the exposure historical default rate in the exposure historical data.
4. The method of claim 3, wherein based on the macro-prediction model, performing a statistical test on the exposure risk parameter according to the predicted value of the target macro-factor and the historical default rate of the exposure in the historical data of the exposure, comprises:
determining a macroscopic year predicted value according to the macroscopic prediction model and the predicted value of the target macroscopic factor;
based on a Nootn model, obtaining a prospective default rate of the opening according to the macroscopic year predicted value and the historical default rate of the opening;
and carrying out statistical test according to the historical default rate and the prospective default rate of the exposure.
5. The method of claim 4, wherein after obtaining the prospective default rate of the exposure according to the macroscopic year prediction value and the historical default rate of the exposure based on a norton model, the method further comprises:
determining the average default rate of the exposure according to the historical default rate of the debt items in the historical data of the exposure;
determining a macroscopic annual evaluation value according to the open prospective default rate and the average default rate on the basis of a Norton model;
and updating the historical default rate of the uncovered debt items according to the macroscopic annual internal evaluation value based on a Norton model.
6. The method of claim 5, wherein the exposure comprises at least two debts, and wherein determining an average default rate of the exposure according to the historical default rate of the debts in the historical data of the exposure comprises:
selecting target debt from at least two debts included in the exposure according to the credit levels of the at least two debts in the historical data of the exposure;
determining an upper limit value and a lower limit value of the historical default rate of the target debt item according to the historical default rate of the target debt item;
and determining the average default rate of the exposure according to the upper limit value and the lower limit value of the historical default rate of the target debt item.
7. The method of claim 6, wherein determining an average default rate of exposure based on the upper and lower historical default rate values of the target debt item comprises:
and carrying out extreme value processing on the debt items in the exposure according to the upper limit value and the lower limit value of the historical default rate of the target debt items to obtain the average default rate of the exposure.
8. The method of claim 1, after obtaining the historical values of the macroscopic year, the historical values of the target macroscopic factors, and the historical data of the exposure, further comprising:
determining a comparison time point according to the time point to be measured;
determining first to-be-tested data and second to-be-tested data related to the target exposure according to the to-be-tested time point, the comparison time point, the target exposure identifier and the historical data of the exposure; the first data to be tested comprises the amount of the debt items and the amount of debt default contained at the time point to be tested, and the second data to be tested comprises the amount of the debt items and the amount of the debt default contained at the comparison time point;
and carrying out stability index inspection according to the first data to be tested and the second data to be tested.
9. The method of claim 1, after obtaining the historical values of the macroscopic year, the historical values of the target macroscopic factors, and the historical data of the exposure, further comprising:
determining debt default quantity and debt non-default quantity related to the credit rating of the target exposure according to the historical data of the exposure, the target exposure identifier and the duration to be measured;
and carrying out Kolmogorov index inspection according to the debt default quantity and the debt non-default quantity which are related to the credit rating of the target exposure.
10. The method of claim 1, wherein after constructing the macro prediction model according to the historical macro year values and the historical values of the target macro factors, the method further comprises:
evaluating the macro prediction model according to a target evaluation index; wherein the target evaluation index comprises at least one of a significance test index and a goodness-of-fit test index.
11. A data processing apparatus, comprising:
the data acquisition module is used for acquiring the historical value of the macroscopic year, the historical numerical value of the target macroscopic factor and the historical data of the exposure;
the model building module is used for building a macroscopic prediction model according to the macroscopic year historical value and the historical numerical value of the target macroscopic factor;
and the verification result output module is used for verifying the risk parameters of the exposure according to the historical data of the exposure based on the macroscopic prediction model.
12. The apparatus of claim 11, further comprising:
the reference model acquisition module is used for acquiring a reference prediction model provided by a user from a visual interface before acquiring the historical values of the macroscopic year, the historical numerical values of the target macroscopic factors and the historical data of the exposure;
a candidate factor extraction module for extracting at least two candidate macroscopic factors from the reference prediction model;
and the target factor extraction module is used for selecting a target macroscopic factor from the at least two candidate macroscopic factors according to the historical distribution condition of the exposure debt and the historical numerical values of the at least two candidate macroscopic factors.
13. The apparatus of claim 11, wherein the verification result output module comprises:
the factor predicted value confirmation unit is used for determining the predicted value of the target macroscopic factor according to the historical numerical value of the target macroscopic factor;
and the statistical test unit is used for carrying out statistical test on the macroscopic prediction model according to the predicted value of the target macroscopic factor and the historical default rate of the exposure in the historical data of the exposure.
14. An electronic device, comprising:
one or more processors;
a storage device for storing one or more programs,
when executed by the one or more processors, cause the one or more processors to implement a data processing method as claimed in any one of claims 1-10.
15. A computer-readable storage medium, on which a computer program is stored, which, when being executed by a processor, carries out the data processing method of any one of claims 1 to 10.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011019016.9A CN112184415A (en) | 2020-09-24 | 2020-09-24 | Data processing method and device, electronic equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011019016.9A CN112184415A (en) | 2020-09-24 | 2020-09-24 | Data processing method and device, electronic equipment and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN112184415A true CN112184415A (en) | 2021-01-05 |
Family
ID=73943402
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011019016.9A Pending CN112184415A (en) | 2020-09-24 | 2020-09-24 | Data processing method and device, electronic equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112184415A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112819608A (en) * | 2021-02-05 | 2021-05-18 | 建信金融科技有限责任公司 | Regional credit prediction method and system based on multiple regression and time series |
| CN114004491A (en) * | 2021-10-29 | 2022-02-01 | 中国建设银行股份有限公司 | Retail risk exposure pool dividing method and device, computer equipment and medium |
-
2020
- 2020-09-24 CN CN202011019016.9A patent/CN112184415A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112819608A (en) * | 2021-02-05 | 2021-05-18 | 建信金融科技有限责任公司 | Regional credit prediction method and system based on multiple regression and time series |
| CN114004491A (en) * | 2021-10-29 | 2022-02-01 | 中国建设银行股份有限公司 | Retail risk exposure pool dividing method and device, computer equipment and medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110246031A (en) | Appraisal procedure, system, equipment and the storage medium of business standing | |
| CN112734559B (en) | Enterprise credit risk evaluation method and device and electronic equipment | |
| CN112184415A (en) | Data processing method and device, electronic equipment and storage medium | |
| CN110688536A (en) | Label prediction method, device, equipment and storage medium | |
| CN107798615A (en) | Declaration form renews charge difficulty Forecasting Methodology and device | |
| CN111090833A (en) | Data processing method, system and related equipment | |
| CN113342939A (en) | Data quality monitoring method and device and related equipment | |
| Amarullah et al. | Planning decision support system using building mall AHP (Analytical Hierarchy Process) | |
| CN113742248A (en) | Method and system for predicting organization process based on project measurement data | |
| CN113793057A (en) | Building bidding and tendering data generation method based on regression analysis model | |
| CN114493208A (en) | Method and device for evaluating engineering project full life cycle, electronic equipment and medium | |
| CN112613983B (en) | Feature screening method and device in machine modeling process and electronic equipment | |
| CN113837799A (en) | Intelligent business site selection method, system, equipment and readable storage medium | |
| JP6192432B2 (en) | Risk weighing system | |
| CN110852322B (en) | Method and device for determining region of interest | |
| CN108446213A (en) | A kind of static code mass analysis method and device | |
| CN112598326A (en) | Model iteration method and device, electronic equipment and storage medium | |
| CN115630708A (en) | Model updating method and device, electronic equipment, storage medium and product | |
| CN115204501A (en) | Enterprise evaluation method and device, computer equipment and storage medium | |
| CN112950362A (en) | Method and device for risk early warning of loan officials, computer equipment and storage medium | |
| CN112163964A (en) | Risk prediction method and device, electronic equipment and storage medium | |
| KR102420952B1 (en) | Loan expansion hypothesis testing system using artificial intelligence and method using the same | |
| EP3794523A1 (en) | System and method for processing distorted or inaccurate input data obtained from an environment | |
| Hai et al. | A New Approach to Calibrating Functional Complexity Weight in Software Development Effort Estimation. Computers 2022, 11, 15 | |
| CN114334137A (en) | Data processing method and device |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |