CN116228403A - Personal bad asset valuation method and system based on machine learning algorithm - Google Patents
Personal bad asset valuation method and system based on machine learning algorithm Download PDFInfo
- Publication number
- CN116228403A CN116228403A CN202310238439.7A CN202310238439A CN116228403A CN 116228403 A CN116228403 A CN 116228403A CN 202310238439 A CN202310238439 A CN 202310238439A CN 116228403 A CN116228403 A CN 116228403A
- Authority
- CN
- China
- Prior art keywords
- data
- personal
- asset
- training
- bad
- 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
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
- Y02P90/30—Computing systems specially adapted for manufacturing
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Software Systems (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- Medical Informatics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Physics & Mathematics (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Mathematical Physics (AREA)
- Artificial Intelligence (AREA)
- Financial Or Insurance-Related Operations Such As Payment And Settlement (AREA)
Abstract
The invention relates to the technical field of financial credit, and particularly discloses a personal bad asset valuation method based on a machine learning algorithm, which comprises the following steps: acquiring current personal credit data of a target user; constructing a bad asset valuation model; training the bad asset valuation model to obtain a trained bad asset valuation model; and inputting the current personal credit data of the target user into the trained bad asset valuation model for valuation so as to output the personal bad asset valuation result of the target user. The invention also discloses a personal bad asset valuation system based on the machine learning algorithm. The personal bad asset estimation method based on the machine learning algorithm can effectively overcome the defects of the existing estimation technology in the current market, and estimation and scoring of the personal credit bad asset can be completed while the accuracy and the efficiency are considered.
Description
Technical Field
The invention relates to the technical field of financial credit, in particular to a personal bad asset valuation method based on a machine learning algorithm and a personal bad asset valuation system based on the machine learning algorithm.
Background
Bad assets are assets that a borrower cannot pay out an interest on time, at full amount, or that a credit business cannot handle in the financial credit field, such as a mortgage property that cannot be handled. To reduce the economic loss caused by bad assets, enterprises in the financial credit-related arts will use big data methods to estimate the asset value of bad asset data, thereby accurately customizing the revenue generation program. When a bad asset is present in an asset of a financial institution such as a bank, how to evaluate the bad asset, i.e., how to determine the recovery price of the bad asset, is a technical problem that needs to be solved by those skilled in the art.
Currently, there is no more sophisticated solution for bad asset valuation for individuals. The existing schemes mainly have two kinds: one is the way in which pricing is done offline. In particular, the bad assets are typically priced by personal experience based on the results of a field survey of the bad assets. However, off-line pricing is inefficient and the pricing results are overly dependent on the personal attributes of the asset security manager, with greater volatility. Secondly, a big data analysis pricing mode is used, according to the mechanism repayment data condition in a period of time in the past, repayment willingness and probability of a customer are presumed, although the efficiency problem is partially solved, accuracy analysis of a specific scene is lacking, repayment results cannot be continuously and effectively monitored and corrected, and follow-up processing after bad asset estimation is lack of attention.
Disclosure of Invention
The invention aims at solving at least one of the technical problems in the prior art, provides a personal bad asset estimation method and a system based on a machine learning algorithm, can effectively overcome the defects of the existing estimation technology in the current market, and can finish estimation and scoring of personal credit bad assets while considering accuracy and efficiency.
As a first aspect of the present invention, there is provided a personal bad asset estimation method based on a machine learning algorithm, comprising the steps of:
step S1: acquiring current personal credit data of a target user;
step S2: constructing a bad asset valuation model;
step S3: training the bad asset valuation model to obtain a trained bad asset valuation model;
step S4: and inputting the current personal credit data of the target user into the trained bad asset valuation model for valuation so as to output the personal bad asset valuation result of the target user.
Further, the building of the bad asset valuation model includes:
constructing a plurality of basic classifiers through a machine learning method;
and constructing the bad asset valuation model according to the plurality of basic classifiers.
Further, the training the bad asset valuation model to obtain a trained bad asset valuation model includes:
three training data for training the bad asset valuation model are obtained, and the weight value of each training data is calculated;
and training the bad asset valuation model according to the three training data and the corresponding weight values thereof to obtain the trained bad asset valuation model.
Further, the three training data are personal credit training data, financial institution training data and macro economy training data respectively;
the personal credit training data comprises personal holding asset data, personal historical credit and repayment data and communication data between the AI collecting robot and a client;
the financial institution training data comprises financial institution same batch all product credit data and financial institution same product or similar product historical credit data;
the macro economic training data includes economic and financial forecast data for a financial rating institution.
Further, the calculating the weight value of each training data further includes:
classifying, grading and scoring the clients according to the personal credit training data of the clients, and comprehensively calculating a weight value alpha of the personal credit training data by referring to the manual assessment;
based on financial institution data provided by a sales room of bad asset packages, training mass data of the historical credit data of all products in the same batch of the financial institution and the same product or similar products of the financial institution along with time domain and macro economy, and obtaining a weight value beta of the financial institution training data;
and collecting economic and financial forecast data of a financial rating institution in the past year, wherein the economic and financial forecast data comprises GDP growth rate, purchasing manager index PMI, consumer price index CPI and commodity expansion rate, and carrying out targeted parameter training and correction by combining personal credit training data in a time domain to obtain a weight value gamma of macroscopic economic training data.
Further, the training the bad asset valuation model according to the three training data and the weight values corresponding to the three training data to obtain the trained bad asset valuation model, and the method further includes:
according to the three training data and the weight values corresponding to the three training data, training the weight and the accuracy of each basic classifier to obtain the trained bad asset estimation model;
taking the clear and refund data fed back in real time as a test set, and preliminarily verifying the accuracy of the trained bad asset estimation model; and delivering the bad asset valuation result obtained during training to an expert consultant for auditing, and finally verifying the accuracy of the trained bad asset valuation model.
Further, the calculation formula of the trained bad asset valuation model H (x) is as follows:
wherein hi (x) is an estimated model function trained by the ith basic classifier, w i Weighting the ith base classifier, T being the baseTotal number of classifiers.
Further, after the current personal credit data of the target user is input into the trained bad asset estimation model to carry out estimation so as to output the personal bad asset estimation result of the target user, the method further comprises the following steps:
and after the personal bad asset real data of the target user are processed into a unified standard format, the unified standard format is used as a newly added training set, and the current bad asset valuation model is continuously optimized and updated.
Further, the trained bad asset valuation models include a net recovery valuation model, a cash flow prediction model, and a stress test and risk control model.
As a second aspect of the present invention, there is provided a machine learning algorithm-based personal bad asset estimation system for implementing the machine learning algorithm-based personal bad asset estimation method described above, the machine learning algorithm-based personal bad asset estimation system comprising:
the acquisition module is used for acquiring the current personal credit data of the target user;
the construction module is used for constructing a bad asset estimation model;
the training module is used for training the bad asset valuation model to obtain a trained bad asset valuation model;
and the valuation module is used for inputting the current personal credit data of the target user into the trained bad asset valuation model to perform valuation so as to output the personal bad asset valuation result of the target user.
The personal bad asset valuation method based on the machine learning algorithm has the following beneficial effects:
(1) High efficiency and high reliability. The asset valuation model based on the financial knowledge graph in the bad asset field reduces the one-by-one analysis of each specific side effect and adverse reaction, but replaces the model with overall personal credit data, financial institution big data and macroscopic economic data, reduces one-by-one judgment of off-line valuation, and the time of transverse comparison and standard format among different characteristic image clients, can rapidly and accurately describe the asset condition of the bad asset package from the whole angle, and is convenient for users and decision makers to quantitatively and transversely compare; the standardized comprehensive model can solve the problem of data backtracking of single logic, ensure accuracy and high efficiency of checking the authenticity of data, and improve analysis efficiency;
(2) The standardization capability is strong. Compared with single-logic big data analysis, the asset valuation model based on the financial knowledge graph in the bad asset field has stronger capacity of carrying out full-flow association analysis on each specific data, and can construct a localized financial knowledge graph instead of completely relying on the data provided by a financial institution selling asset packages. When the newly collected and provided data formats and contents and the original logic have certain access, the data formats and contents can be processed through a preset data standardization flow, the whole evaluation estimation system is not affected, all other links can not be linked, and the maintenance cost of the estimation system is reduced;
(3) The effectiveness is high, and the repeatability is strong. The asset valuation model based on the financial knowledge graph in the bad asset field not only trains for individual cases or specific links, but also covers the whole process from credit generation to clearing and subsequent data tracking, and the repeatability and the reliability are limited by the level of financial professionals, accepted data defects, errors and the like and are much smaller than those of the prior art; in addition, the invention fully considers the great financial environment, including the important influence caused by development expectation and policy change, and takes the important influence as training parameter feedback in the training of the asset valuation model, which is an original technical scheme not considered by various other valuation methods in the existing market;
(4) The iterative optimization space is large. In the process of advancing and continuously optimizing the evaluation method, when new collected data or new research results are defined with the original logic, the evaluation method does not need to redesign the whole evaluation logic structure (such as manually examining part of conclusions, adding and deleting the existing logic and architecture), only needs to apply various machine learning algorithms covered by the related integrated learning to retrain parameters, and saves the cost of optimizing iteration.
Drawings
The accompanying drawings are included to provide a further understanding of the invention, and are incorporated in and constitute a part of this specification, illustrate the invention and together with the description serve to explain, without limitation, the invention.
FIG. 1 is a flow chart of a method for valuation of a personal undesirable asset based on a machine learning algorithm provided by the present invention.
FIG. 2 is a training flow chart of the bad asset valuation model provided by the present invention.
FIG. 3 is a classification diagram of a trained bad asset valuation model provided by the invention.
Detailed Description
In order to further describe the technical means and effects adopted by the invention to achieve the preset aim, the following detailed description refers to the specific implementation, structure, characteristics and effects of a method for estimating a personal bad asset based on a machine learning algorithm according to the invention, which are described in detail below with reference to the accompanying drawings and the preferred embodiments. It will be apparent that the described embodiments are some, but not all, embodiments of the invention. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to fall within the scope of the invention.
It should be noted that the terms "first," "second," and the like in the description and the claims of the present invention and the above figures are used for distinguishing between similar objects and not necessarily for describing a particular sequential or chronological order. It is to be understood that the data so used may be interchanged where appropriate in order to describe the embodiments of the invention herein. Furthermore, the terms "comprises," "comprising," and "having," and any variations thereof, are intended to cover a non-exclusive inclusion, such that a process, method, system, article, or apparatus that comprises a list of steps or elements is not necessarily limited to those steps or elements expressly listed but may include other steps or elements not expressly listed or inherent to such process, method, article, or apparatus.
In this embodiment, a method for estimating a personal bad asset based on a machine learning algorithm is provided, as shown in fig. 1, and includes the following steps:
step S1: acquiring current personal credit data of a target user;
step S2: constructing a bad asset valuation model;
preferably, the building of the bad asset valuation model comprises:
constructing a plurality of basic classifiers through a machine learning method;
and constructing the bad asset valuation model according to the plurality of basic classifiers.
Step S3: training the bad asset valuation model to obtain a trained bad asset valuation model;
preferably, as shown in fig. 2, the training the bad asset valuation model to obtain a trained bad asset valuation model includes:
three training data for training the bad asset valuation model are obtained, and the weight value of each training data is calculated;
and training the bad asset valuation model according to the three training data and the corresponding weight values thereof to obtain the trained bad asset valuation model.
Specifically, the three training data are personal credit training data, financial institution training data and macro economy training data respectively;
the personal credit training data comprises personal holding asset data, personal historical credit and repayment data and communication data between the AI collecting robot and a client;
the financial institution training data comprises financial institution same batch all product credit data and financial institution same product or similar product historical credit data;
the macro economic training data includes economic and financial forecast data for a financial rating institution.
Specifically, the calculating the weight value of each training data further includes:
based on the own financial database of the company and the knowledge graph of the bad asset financial field constructed by the knowledge graph, sorting and scoring the data of the personal holding asset (the cash flow data of the asset and the individual or related account under the held personal name), the historical credit and repayment data (the repayment and overdue data of the similar or similar credit products in the past), the communication between the AI collecting robot and the client (the negotiation repayment dialogue of the client, including the exhibition expectation, the interest expectation and the like), and the like, classifying, grading and scoring the client according to the repayment intention of the bad asset of the client and the repayment capability and the like of the current holding asset, referring to the professional institution and expert committee comments of the corresponding financial field, and comprehensively calculating the weight value alpha of the personal credit training data;
based on financial institution data provided by a sales room of bad asset packages, training mass data of the historical credit data of all products in the same batch of the financial institution and the same product or similar products of the financial institution along with time domain and macro economy, and obtaining a weight value beta of the financial institution training data;
and collecting economic and financial forecast data of a financial rating institution over the years, including but not limited to GDP growth rate (forecast and actual values), purchasing manager index PMI, consumer price index CPI, commodity expansion rate and the like, and carrying out targeted parameter training and correction by combining personal credit training data of a time domain to obtain a weight value gamma of macro economic training data.
Specifically, the training the bad asset valuation model according to the three training data and the weight values corresponding to the three training data to obtain the trained bad asset valuation model, and the method further includes:
according to the three training data and the weight values corresponding to the three training data, training the weight and the accuracy of each basic classifier to obtain the trained bad asset estimation model; wherein each basic classifier is integrated by using a weighted average method, and the calculation formula of the trained bad asset estimation model H (x) is as follows:
wherein hi (x) is an estimated model function trained by the ith basic classifier, w i And (3) weighting the ith basic classifier, wherein T is the total number of the basic classifiers. In addition, in the case of the optical fiber,
it should be noted that each basic classifier uses a different machine learning modeling method, for example, linear regression, random forest, SVM support vector machine, etc.
The method is characterized in that the clearing and withdrawing data fed back in real time are used as a test set, and the accuracy of the trained bad asset estimation model is preliminarily verified; and delivering the bad asset valuation result (partial financial risk early warning data) obtained during training to an expert consultant for auditing, and finally verifying the accuracy of the trained bad asset valuation model.
Step S4: and inputting the current personal credit data of the target user into the trained bad asset valuation model for valuation so as to output the personal bad asset valuation result of the target user.
Further, after the current personal credit data of the target user is input into the trained bad asset estimation model to carry out estimation so as to output the personal bad asset estimation result of the target user, the method further comprises the following steps:
with the use of products and continuous updating iteration, the real data of the personal bad asset of the target user are processed into a unified standard format, and then are used as a new training set to continuously optimize and update the current bad asset estimation model, so that the comprehensive accuracy and the effectiveness of the model are improved in an iteration mode.
Further, the bad asset valuation model trained through the steps belongs to a basic model template, and a sub-model classification training interface aiming at different specific requirements is reserved in advance, and the bad asset valuation model mainly comprises the following three sub-modules: as shown in FIG. 3, the trained bad asset valuation models include a net recovery valuation model, a cash flow prediction model, and a stress test and risk control model.
Specifically, (1) the net recovery valuation model is the base sub-model of the bad asset valuation model, considering only the cleanable value of the bad asset, i.e., the proportion of the asset pack discount in the ideal case. (2) The cash flow prediction model is based on a net recovery estimation model, increases the estimation weight of a time domain, and performs comprehensive evaluation calculation on the clearing and refunding efficiency and account period time, so that cash flow change during bad credit processing can be effectively reflected and predicted. (3) The pressure test and risk control model is a sub-model which combines the risk of the bad asset financial knowledge graph and the training of fraud data in a pertinence manner, the estimated weight of risk management is increased, the integral risk of the bad asset package, the personal repayment willingness of individual clients, the bad account possibility and the like can be comprehensively evaluated, the dialogue data of the communication robot is cleared and collected in combination with AI, the pertinence analysis is obtained by mapping the comprehensive risk to the discount price and cash flow obstacle of the asset estimated value, the scene related by the financial knowledge graph is provided, and the solution is correspondingly designed.
In the embodiment of the invention, firstly, according to the standardized requirement of the financial knowledge graph, all relevant data are arranged according to a unified standard, and then, key parameters are extracted in three directions according to the data: (1) personal credit data weight alpha (2) financial institution data weight beta (3) macro economic data weight gamma, and generating a basic model based on the personal credit data weight alpha (2) financial institution data weight beta (3) macro economic data weight gamma; performing iterative training through machine learning algorithms such as ensemble learning, and continuously adding new clinical feedback data along with the practice of a product system to continuously optimize a bad asset estimation model; wherein the formula for the discount rate estimate Index for the individual undesirable asset is as follows: index estimate = α x F (β, γ), the equation based net recovery estimate discount rate, and other composite costs such as cash recovery cycles need to be considered in the module.
As another embodiment of the present invention, there is provided a machine learning algorithm-based personal bad asset estimation system for implementing the machine learning algorithm-based personal bad asset estimation method described above, the machine learning algorithm-based personal bad asset estimation system comprising:
the acquisition module is used for acquiring the current personal credit data of the target user;
the construction module is used for constructing a bad asset estimation model;
the training module is used for training the bad asset valuation model to obtain a trained bad asset valuation model;
and the valuation module is used for inputting the current personal credit data of the target user into the trained bad asset valuation model to perform valuation so as to output the personal bad asset valuation result of the target user.
In particular, the invention relates to a personal loan bad property valuation system based on a financial knowledge graph in the bad credit field, an integrated learning algorithm and other machine learning algorithms.
It should be noted that the financial knowledge graph of the present invention relates to the following functional modules and fields, as shown in table 1 below:
TABLE 1
Project management: and (3) performing visual display and management operation aiming at a plurality of bad asset packs of different financial institutions, different periods and different scales, and constructing a unified and standardized project management system.
Asset pack management: based on the clear segmentation of the project, various data carried by the bad asset package are classified into personal credit data, institution financial data and macroscopic economic data (macroscopic economic data sources are various and are not necessarily bound with the asset package data), and are cleaned, characteristic engineering, split and integrated in a targeted manner, and various standards are precisely classified according to the specifications of the knowledge graph, so that a basis is laid for the subsequent differentiation and customization treatment, and the barriers to standardized systems and data management caused by the great difference between the sources of the bad asset and the metadata formats are avoided.
End-of-job investigation and end-of-job results: in the field of the disposal of bad assets, false data and malicious fraud have been important factors constituting a systematic financial risk. In the module, the invention designs a risk control model based on self financial knowledge graph through the personal credit related data of the whole process, classifies and pertinently processes clients with risk scores higher than a standard line, provides follow-up information in time and corrects an estimated model.
Estimation model: the valuation model is one of the technical core modules of the invention, and the content of the valuation model comprises a plurality of sub-modules and built-in content such as valuation model pre-training, net recycling valuation model analysis, cash flow model analysis, pressure test and subsequent data tracking.
In summary, the invention relates to a bad asset comprehensive evaluation and scoring system based on a financial knowledge graph in the field of bad assets, which uses a visualization technology to describe knowledge resources and carriers thereof in the related field, and excavates, analyzes, builds, draws and displays knowledge and interrelations between the knowledge resources and carriers. On the basis, the evaluation and grading of the personal credit bad assets are carried out by a machine learning method such as ensemble learning and the like through the high-standard data collected and processed by the personal credit bad assets; meanwhile, the continuous tracking of bad asset data entering the clearing time is kept, and the refund efficiency and the refund period of the asset pack are corrected in real time based on the continuous tracking. The method can effectively make up the defects of the existing valuation technology in the current market, and finish the valuation and the grading of the full life cycle from the borrowing generation to the credit overdue of the personal credit bad asset to the clear period and the personal credit grading and reconstruction while considering the accuracy and the efficiency.
The present invention is not limited to the above-mentioned embodiments, but is intended to be limited to the following embodiments, and any modifications, equivalents and modifications can be made to the above-mentioned embodiments without departing from the scope of the invention.
Claims (10)
1. A method for valuating a personal undesirable asset based on a machine learning algorithm, comprising the steps of:
step S1: acquiring current personal credit data of a target user;
step S2: constructing a bad asset valuation model;
step S3: training the bad asset valuation model to obtain a trained bad asset valuation model;
step S4: and inputting the current personal credit data of the target user into the trained bad asset valuation model for valuation so as to output the personal bad asset valuation result of the target user.
2. The method for valuation of a personal asset based on a machine learning algorithm of claim 1, wherein the constructing the valuation model of the asset comprises:
constructing a plurality of basic classifiers through a machine learning method;
and constructing the bad asset valuation model according to the plurality of basic classifiers.
3. The method of claim 2, wherein training the asset valuation model to obtain a trained asset valuation model comprises:
three training data for training the bad asset valuation model are obtained, and the weight value of each training data is calculated;
and training the bad asset valuation model according to the three training data and the corresponding weight values thereof to obtain the trained bad asset valuation model.
4. A method of evaluating a personal asset based on a machine learning algorithm as defined in claim 3, wherein the three training data are personal credit training data, financial institution training data, macro economic training data, respectively;
the personal credit training data comprises personal holding asset data, personal historical credit and repayment data and communication data between the AI collecting robot and a client;
the financial institution training data comprises financial institution same batch all product credit data and financial institution same product or similar product historical credit data;
the macro economic training data includes economic and financial forecast data for a financial rating institution.
5. The method of estimating a personal undesirable asset based on a machine learning algorithm of claim 4 wherein said calculating a weight value for each training data further comprises:
classifying, grading and scoring the clients according to the personal credit training data of the clients, and comprehensively calculating a weight value alpha of the personal credit training data by referring to the manual assessment;
based on financial institution data provided by a sales room of bad asset packages, training mass data of the historical credit data of all products in the same batch of the financial institution and the same product or similar products of the financial institution along with time domain and macro economy, and obtaining a weight value beta of the financial institution training data;
and collecting economic and financial forecast data of a financial rating institution in the past year, wherein the economic and financial forecast data comprises GDP growth rate, purchasing manager index PMI, consumer price index CPI and commodity expansion rate, and carrying out targeted parameter training and correction by combining personal credit training data in a time domain to obtain a weight value gamma of macroscopic economic training data.
6. The method for valuating a personal asset based on a machine learning algorithm of claim 3, wherein the training the valuation model of the asset according to three training data and their corresponding weights to obtain the trained valuation model of the asset further comprises:
according to the three training data and the weight values corresponding to the three training data, training the weight and the accuracy of each basic classifier to obtain the trained bad asset estimation model;
taking the clear and refund data fed back in real time as a test set, and preliminarily verifying the accuracy of the trained bad asset estimation model; and delivering the bad asset valuation result obtained during training to an expert consultant for auditing, and finally verifying the accuracy of the trained bad asset valuation model.
7. The method for valuation of a personal asset based on a machine learning algorithm of claim 6, wherein the trained valuation model H (x) is calculated as:
wherein hi (x) is an estimated model function trained by the ith basic classifier, w i And (3) weighting the ith basic classifier, wherein T is the total number of the basic classifiers.
8. The method for valuating a personal asset based on a machine learning algorithm of claim 1, wherein after inputting the current personal credit data of the target user into the trained valuation model for valuation to output the personal valuation result of the target user, further comprising:
and after the personal bad asset real data of the target user are processed into a unified standard format, the unified standard format is used as a newly added training set, and the current bad asset valuation model is continuously optimized and updated.
9. The method of claim 1, wherein the trained asset valuation model comprises a net recovery valuation model, a cash flow prediction model, and a stress test and risk control model.
10. A machine learning algorithm-based personal undesirable asset valuation system for implementing the machine learning algorithm-based personal undesirable asset valuation method of any of claims 1-9, the machine learning algorithm-based personal undesirable asset valuation system comprising:
the acquisition module is used for acquiring the current personal credit data of the target user;
the construction module is used for constructing a bad asset estimation model;
the training module is used for training the bad asset valuation model to obtain a trained bad asset valuation model;
and the valuation module is used for inputting the current personal credit data of the target user into the trained bad asset valuation model to perform valuation so as to output the personal bad asset valuation result of the target user.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310238439.7A CN116228403A (en) | 2023-03-13 | 2023-03-13 | Personal bad asset valuation method and system based on machine learning algorithm |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310238439.7A CN116228403A (en) | 2023-03-13 | 2023-03-13 | Personal bad asset valuation method and system based on machine learning algorithm |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN116228403A true CN116228403A (en) | 2023-06-06 |
Family
ID=86574933
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310238439.7A Pending CN116228403A (en) | 2023-03-13 | 2023-03-13 | Personal bad asset valuation method and system based on machine learning algorithm |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116228403A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117217807A (en) * | 2023-11-08 | 2023-12-12 | 四川智筹科技有限公司 | Bad asset valuation algorithm based on multi-mode high-dimensional characteristics |
-
2023
- 2023-03-13 CN CN202310238439.7A patent/CN116228403A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117217807A (en) * | 2023-11-08 | 2023-12-12 | 四川智筹科技有限公司 | Bad asset valuation algorithm based on multi-mode high-dimensional characteristics |
| CN117217807B (en) * | 2023-11-08 | 2024-01-26 | 四川智筹科技有限公司 | Bad asset estimation method based on multi-mode high-dimensional characteristics |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Kara et al. | A data mining-based framework for supply chain risk management | |
| Ogwueleka et al. | Neural network and classification approach in identifying customer behavior in the banking sector: A case study of an international bank | |
| TW580627B (en) | System and method for efficiently providing due diligence knowledge and a computer therefor | |
| Smith et al. | Neural networks in business: techniques and applications | |
| Bose et al. | Big data, data analytics and artificial intelligence in accounting: An overview | |
| TW530234B (en) | Methods and systems for efficiently sampling portfolios for optimal underwriting | |
| KR102009309B1 (en) | Management automation system for financial products and management automation method using the same | |
| Wu et al. | The analysis of credit risks in agricultural supply chain finance assessment model based on genetic algorithm and backpropagation neural network | |
| JP2004500646A (en) | Competitive bidding revenue simulation method and system | |
| WO2011133904A1 (en) | Performance analytics based on high performance indices | |
| JP2004500641A (en) | Method and system for automatically estimating credit score evaluation value | |
| JP2004500642A (en) | Methods and systems for assessing cash flow recovery and risk | |
| Wang et al. | Integrating decision tree with back propagation network to conduct business diagnosis and performance simulation for solar companies | |
| CN107133862A (en) | Dynamic produces the method and system of the detailed transaction payment experience of enhancing credit evaluation | |
| CN116228403A (en) | Personal bad asset valuation method and system based on machine learning algorithm | |
| Wei | A machine learning algorithm for supplier credit risk assessment based on supply chain management | |
| Li et al. | Tracking down financial statement fraud by analyzing the supplier-customer relationship network | |
| Dai et al. | Audit analytics: A field study of credit card after-sale service problem detection at a major bank | |
| Biswas et al. | Automated credit assessment framework using ETL process and machine learning | |
| CN115619571A (en) | Financing planning method, system and device | |
| KR102499182B1 (en) | Loan regular auditing system using artificia intellicence | |
| Ofoegbu | A comparison of four machine learning algorithms to predict product sales in a retail store | |
| Wanniarachchi et al. | Macroeconomic event base expert advisor for forex trades: through algo trading | |
| KR102499183B1 (en) | Loan regular auditing system using artificia intellicence | |
| Pathak et al. | Impact of Artificial Intelligence: Applications, Transformation Strategy and Future Potential |
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 |