CN118799058B - A loan risk assessment method and system based on artificial intelligence - Google Patents
A loan risk assessment method and system based on artificial intelligence Download PDFInfo
- Publication number
- CN118799058B CN118799058B CN202410779852.9A CN202410779852A CN118799058B CN 118799058 B CN118799058 B CN 118799058B CN 202410779852 A CN202410779852 A CN 202410779852A CN 118799058 B CN118799058 B CN 118799058B
- Authority
- CN
- China
- Prior art keywords
- risk
- loan
- financial
- market
- enterprise
- 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.)
- Active
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; 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
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/044—Recurrent networks, e.g. Hopfield networks
- G06N3/0442—Recurrent networks, e.g. Hopfield networks characterised by memory or gating, e.g. long short-term memory [LSTM] or gated recurrent units [GRU]
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- Finance (AREA)
- General Engineering & Computer Science (AREA)
- Accounting & Taxation (AREA)
- Evolutionary Computation (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Bioinformatics & Computational Biology (AREA)
- Evolutionary Biology (AREA)
- Biomedical Technology (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Molecular Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Computational Linguistics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Computing Systems (AREA)
- General Health & Medical Sciences (AREA)
- Biophysics (AREA)
- Health & Medical Sciences (AREA)
- Development Economics (AREA)
- Economics (AREA)
- Marketing (AREA)
- Strategic Management (AREA)
- Technology Law (AREA)
- General Business, Economics & Management (AREA)
- Financial Or Insurance-Related Operations Such As Payment And Settlement (AREA)
Abstract
The invention relates to the technical field of credit risk, in particular to a loan risk assessment method and a loan risk assessment system based on artificial intelligence, comprising the following steps, and collecting historical loan records, financial reports, bank flowing water and market behavior data of enterprises, integrating the collected data, analyzing the fund flowing and debt repaying capacity, evaluating the financial conditions of the enterprises and generating a financial health index. According to the invention, the precision and efficiency of preliminary credit examination are improved by introducing the financial health index, the consumption of resources is reduced, the influence of market fluctuation on enterprise credit can be reflected in real time by dynamically analyzing the relevance of loan behaviors and market change, more prospective risk assessment is provided, the application of product matching index information optimizes the loan product recommendation flow, and the financial products are ensured to be highly consistent with enterprise demands, so that the customer satisfaction degree and the effective delivery of the loan products are improved, and the accuracy of risk level classification enables loan decisions to be more scientific.
Description
Technical Field
The invention relates to the technical field of credit risk, in particular to a loan risk assessment method and system based on artificial intelligence.
Background
The field of credit risk technology relates to the use of different methods and models to predict the likelihood of loan violations and their financial consequences. In the field, financial institutions and other lending entities use various statistical, machine learning, and artificial intelligence algorithms to evaluate borrower credit status, including credit history, repayment capabilities, and economic conditions, to help financial institutions make more informed decisions when approving loans and pricing conditions, thereby reducing default rates and potential financial losses.
The loan risk assessment method is a systematic process for assessing the default risk of borrowers by analyzing the credit records, financial status, repayment capability and other relevant factors of the borrowers to determine the credit rating thereof, and is mainly used for assessing the potential risk before loan release to determine whether to approve the loan and determine the proper loan condition and interest rate.
The traditional method has the defects that the real-time data processing and dynamic risk prediction are insufficient, the immediate change of the market cannot be effectively reflected, the risk of misjudgment is increased, the lack of a flexible loan product matching mechanism is also a great defect of the traditional method, in the traditional evaluation flow, the product recommendation fails to fully consider the requirements and the latest financial conditions of enterprises, the resource allocation efficiency is low, the idle or excessive liability of funds is caused, the traditional method has limited capability in the aspects of loan structure optimization and risk level dynamic adjustment, the flexibility and the initiative of financial institutions in strategy adjustment and risk management are limited, the response to economic changes is not timely, and the risk of financial loss is increased.
Disclosure of Invention
The invention aims to solve the defects in the prior art, and provides an artificial intelligence-based loan risk assessment method and system.
In order to achieve the purpose, the invention adopts the following technical scheme that the loan risk assessment method based on artificial intelligence comprises the following steps:
S1, collecting historical loan records, financial reports, bank flowing water and market behavior data of enterprises, integrating the collected data, analyzing the flowability and repaying capacity of funds, evaluating the financial conditions of the enterprises and generating financial health indexes;
S2, based on the financial health index, extracting credit risk factors, quantitatively analyzing the relevance of lending behaviors and market changes, and calculating the default probability of an enterprise through the risk factors to obtain default probability scores;
s3, identifying loan products with optimal matching degree by utilizing a long-short-term memory network and a weighted Markov distance algorithm based on the default probability score, and analyzing the corresponding relation between product conditions and enterprise requirements to obtain product matching index information;
S4, based on the product matching index information, adopting a decision tree algorithm and a gating circulation unit to perform correlation analysis of credit levels and default risks, and quantifying risk parameters for each credit level to obtain a risk level classification result;
S5, based on the risk level classification result, evaluating the current loan structure, and identifying potential financing gaps and optimization points through industry comparison analysis to generate a loan structure optimization record;
And S6, based on the loan structure optimization record, carrying out dynamic risk assessment by utilizing the historical data of financial performance, and predicting the variation trend of the liability to obtain a loan risk dynamic monitoring result.
The invention improves, the said financial health index includes financial structural index, repayment ability index, operational ability index, profitability index, development ability index, the said breaking probability score includes credit score, financial condition change information, market behavior analysis result, the said product matches index information and includes loan kind, guarantee measure, loan deadline, interest rate option, the said risk classification result includes breaking risk rating, expected loss rate, credit recovery potential, the said loan structural optimization record includes adjusting loan proportion, risk dispersion tactics, debt reorganization plan, the said loan risk dynamic monitoring result includes the breaking probability prediction after adjustment, market fluctuation response, risk relieving measure updating result.
The invention is improved by collecting historical loan records, financial reports, bank running water and market behavior data of enterprises, integrating the collected data, analyzing the mobility and repayment capacity of funds, evaluating the financial condition of the enterprises, and generating financial health indexes, wherein the steps are as follows:
S101, collecting historical loan records, financial reports, bank running water and market behavior data of enterprises, and cleaning data, including consistency checking, removing repeated items, correcting error formats and filling missing values, to generate an integrated data set;
S102, analyzing cash inflow and outflow modes of enterprises based on the integrated data set, evaluating the fluctuation of the cash flow of month and year, and determining the stability and repayment capability of the cash flow to obtain a cash flow evaluation result;
And S103, based on the fund liquidity assessment result, assessing the financial health state of the enterprise, and calculating key financial ratios, including the balance ratio and the yield, by adopting an industry standardized ratio to obtain a financial health index.
The improvement of the invention is that based on the financial health index, the extraction of credit risk factors is executed, the relevance of lending behavior and market change is quantitatively analyzed, the default probability of enterprises is calculated through the risk factors, and the step of obtaining the default probability score is specifically as follows:
s201, identifying risk factors associated with enterprise default probabilities based on the financial health index, including debt levels and fund flow fluctuations, collecting factors and constructing a risk analysis framework to generate a risk analysis infrastructure;
S202, analyzing the relevance of lending behaviors and market dynamics based on the risk analysis basic structure, and extracting market sensitivity indexes by identifying the influence of market condition changes on the lending behaviors to obtain a market influence assessment result;
And S203, integrating the risk factors and the market dynamic data by adopting a weight distribution method based on the market influence evaluation result, and calculating the default probability of the enterprise under the current market condition to obtain a default probability score.
The invention is improved in that based on the default probability score, a long-term memory network and a weighted mahalanobis distance algorithm are utilized to identify loan products with optimal matching degree, and the corresponding relation between product conditions and enterprise requirements is analyzed, and the steps of obtaining product matching index information are specifically as follows:
S301, based on the default probability score, screening financial indexes including profit margin and cash flow rate by adopting a long-short-period memory network, comparing the financial indexes with the characteristics of the existing loan products, identifying the matching degree of the loan products and the financial conditions of enterprises, and generating product matching efficiency analysis information;
s302, based on the product matching efficiency analysis information, comparing loan conditions of each product, including repayment deadline and interest rate, with the fund requirements and compensation capabilities of enterprises by using a weighted Markov distance algorithm, and evaluating the matching effectiveness to obtain a loan adaptation degree checking result;
And S303, integrating financial stability scores and growth potential evaluations of enterprises based on the loan adaptation degree checking result, listing adaptation degree and potential risks of each recommended product, and generating product matching index information.
The invention improves that the weighted mahalanobis distance algorithm is as follows:
a weighted Marshall distance D WM between the loan product and the enterprise financial demand is calculated, where x is the target enterprise financial demand vector, y is the loan product feature vector, S is the covariance matrix of the overall enterprise financial demand, and W is the weight matrix.
The improvement of the invention is that based on the product matching index information, a decision tree algorithm and a gating circulation unit are adopted to analyze the relevance of the credit level and the default risk, and the risk parameters are quantified for each credit level, so that the step of obtaining the risk level classification result is as follows:
S401, based on the product matching index information, adopting a decision tree algorithm to collect enterprise historical default data of a plurality of credit grades, analyzing a risk mode of the data, identifying risk factors sensitive to the credit grades, and generating a credit risk factor set;
S402, based on the credit risk factor set, applying a gating circulation unit, setting a risk threshold for each level by comparing and analyzing risk exposure under a plurality of credit levels, and evaluating potential credit variation according to the threshold to obtain a credit risk quantification model;
And S403, classifying the risk grades of the enterprises based on the credit risk quantification model, wherein the grades comprise low risk, medium risk and high risk, and designating matched loan conditions and interest rates for each grade to generate a risk grade classification result.
The invention is improved in that, based on the risk level classification result, the current loan structure is evaluated, potential financing gaps and optimization points are identified through industry comparison analysis, and the step of generating a loan structure optimization record is specifically as follows:
S501, based on the risk level classification result, comparing consistency of enterprise multi-class loan conditions and risk levels, checking coverage and limitation conditions of loan products, identifying areas in excessive risk concentration, and generating a loan structure preliminary analysis record;
s502, based on the preliminary analysis record of the loan structure, using industry average data as a reference, evaluating the deviation degree of the enterprise loan structure and industry average, and determining financing gaps and structure shortages to obtain financing gaps and an optimization scheme;
And S503, based on the financing gap and the optimization scheme, setting optimization measures including loan condition adjustment, new loan product introduction or existing product condition modification, matching the requirements and risk bearing capacity of enterprises, and generating a loan structure optimization record.
The invention improves, based on the loan structure optimization record, the historical data of financial performance is utilized to carry out dynamic risk assessment, the change trend of the liability is predicted, and the steps for obtaining the dynamic monitoring result of the loan risk are as follows:
S601, based on the loan structure optimization record, collecting historical financial data and current market dynamic information of enterprises, analyzing the influence of the data on the enterprise property liability ratio and cash flow, identifying the change trend of financial performance, and generating market and financial dynamic analysis information;
s602, based on the market and financial dynamic analysis information, evaluating the accuracy of the asset liability prediction model, dynamically adjusting risk parameters according to the market, and optimizing the response capability of the model to obtain an updated record of the risk evaluation model;
And S603, updating records based on the risk assessment model, performing risk monitoring, reflecting the changes of market and financial conditions in real time, optimizing the real-time performance of risk early warning and management measures, and generating a loan risk dynamic monitoring result.
An artificial intelligence based loan risk assessment system, the system comprising:
the data integration module integrates historical loan records, financial reports, bank flowing water and market behavior data of enterprises, analyzes the fund flowing and debt repaying capacity, evaluates the financial conditions of the enterprises and generates financial health indexes;
The credit risk assessment module extracts credit risk factors based on the financial health index, quantitatively analyzes the relevance of lending behaviors and market changes, calculates the enterprise default probability, and obtains a default probability score;
The product matching analysis module evaluates the financial stability and growth potential of enterprises based on the default probability score, identifies and matches the optimal loan products, analyzes the corresponding relation between the product conditions and the enterprise demands, and acquires product matching index information;
The risk level classification module analyzes the relevance of the credit level and the default risk by using a decision tree algorithm and a gating circulation unit according to the product matching index information, quantifies the risk parameter of each credit level, and classifies the enterprise risk level to obtain a risk level classification result;
And the loan structure optimization module evaluates the current loan structure based on the risk level classification result, identifies potential financing gaps and optimization points in industry comparison, quantifies defects in the loan structure, updates a risk assessment model and matches market changes, and generates a loan structure optimization record.
Compared with the prior art, the invention has the advantages and positive effects that:
According to the invention, the introduction of the financial health index improves the precision and efficiency of preliminary credit examination, reduces the consumption of resources, can reflect the influence of market fluctuation on enterprise credit in real time by dynamically analyzing the relevance of loan behaviors and market change, provides more prospective risk assessment, optimizes the recommending flow of loan products by applying the product matching index information, ensures that the financial products are highly consistent with the enterprise demands, thereby improving the customer satisfaction degree and the effective delivery of the loan products, ensuring the accuracy of risk level classification to make the loan decision more scientific, further enhancing the efficiency of fund configuration and the initiative of risk management by optimizing the loan structure, and enhancing the adaptability and risk prevention and control capability of financial institutions in continuously changing markets.
Drawings
FIG. 1 is a flow chart of a loan risk assessment method based on artificial intelligence;
FIG. 2 is a schematic diagram showing a detailed flow of step S1 in a loan risk assessment method based on artificial intelligence;
FIG. 3 is a schematic diagram showing a detailed flow of step S2 in the loan risk assessment method based on artificial intelligence;
FIG. 4 is a schematic diagram showing a detailed flow of step S3 in the loan risk assessment method based on artificial intelligence;
FIG. 5 is a schematic diagram showing a detailed process of step S4 in the loan risk assessment method based on artificial intelligence;
FIG. 6 is a schematic diagram showing a detailed process of step S5 in the loan risk assessment method based on artificial intelligence;
FIG. 7 is a schematic diagram showing a detailed flow of step S6 in the loan risk assessment method based on artificial intelligence;
FIG. 8 is a block diagram of an artificial intelligence based loan risk assessment system according to the present invention.
Detailed Description
The present invention will be described in further detail with reference to the drawings and examples, in order to make the objects, technical solutions and advantages of the present invention more apparent. It should be understood that the specific embodiments described herein are for purposes of illustration only and are not intended to limit the scope of the invention.
In the description of the present invention, it should be understood that the terms "length," "width," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," "outer," etc. indicate or are based on the orientation or positional relationship shown in the drawings, merely to facilitate description of the invention and simplify description, and do not indicate or imply that the devices or elements referred to must have a specific orientation, be configured and operated in a specific orientation, and thus should not be construed as limiting the invention. Furthermore, in the description of the present invention, the meaning of "a plurality" is two or more, unless explicitly defined otherwise.
Examples
Referring to FIG. 1, the invention provides a technical scheme that the loan risk assessment method based on artificial intelligence comprises the following steps:
S1, collecting historical loan records, financial reports, bank flowing water and market behavior data of enterprises, integrating the collected data, analyzing the flowability and repaying capacity of funds, evaluating the financial conditions of the enterprises and generating financial health indexes;
S2, based on the financial health index, extracting credit risk factors, quantitatively analyzing the relevance of lending behaviors and market changes, and calculating the default probability of enterprises through the risk factors to obtain default probability scores;
S3, identifying loan products with optimal matching degree by utilizing a long-short-term memory network and a weighted Markov distance algorithm and through financial stability and growth potential of enterprises based on the default probability score, and analyzing the corresponding relation between product conditions and enterprise requirements to obtain product matching index information;
S4, based on the product matching index information, carrying out correlation analysis of credit levels and default risks by adopting a decision tree algorithm and a gating circulation unit, quantifying risk parameters for each credit level, and dividing enterprise risk levels to obtain a risk level classification result;
s5, evaluating the current loan structure based on the risk level classification result, identifying potential financing gaps and optimization points through industry comparison analysis, quantifying insufficient and potential improvement points in the loan structure, and generating a loan structure optimization record;
And S6, based on the loan structure optimization record, carrying out dynamic risk assessment by utilizing the historical data of financial performance, predicting the change trend of the liability, updating a risk assessment model and matching market change to obtain a loan risk dynamic monitoring result.
The financial health index comprises a financial structure index, a repayment capability index, an operational capability index, a profitability index and a development capability index, the offence probability score comprises credit rating, financial condition change information and a market behavior analysis result, the product matching index information comprises loan types, loan deadlines and interest rate options, the risk level classification result comprises offence risk rating, expected loss rate and credit recovery potential, the loan structure optimization record comprises an adjusted loan proportion, a risk dispersion strategy and a debt recombination plan, and the loan risk dynamic monitoring result comprises an adjusted offence probability prediction, a market fluctuation response and a risk relief measure updating result.
Referring to fig. 2, historical loan records, financial reports, banking running water and market behavior data of enterprises are collected, the collected data are integrated, the mobility of funds and the repayment capability are analyzed, the financial condition of the enterprises is evaluated, and the steps for generating the financial health index are specifically as follows:
S101, collecting historical loan records, financial reports, bank running water and market behavior data of enterprises, and cleaning data, including consistency checking, removing repeated items, correcting error formats and filling missing values, to generate an integrated data set;
s101, firstly classifying and sorting historical loan records of enterprises, financial reports, bank running water and market behavior data, carrying out detailed examination on each type of data to ensure the integrity and accuracy of the data, for example, carrying out data integrity examination on the loan records to ensure that no missing internal and external liabilities exist, checking the data consistency of a property liability table and a damage and benefit table in the financial reports to ensure the authenticity and reliability of the report data, carrying out time continuity examination on bank running water to ensure that no missing time period exists, checking and verifying the frequency and volume of transactions on the market behavior data to ensure that the data reflects the real market condition, and merging the data into an integrated data set after the data cleaning is finished.
S102, analyzing cash inflow and outflow modes of enterprises based on the integrated data set, evaluating the fluctuation of the cash flow of month and year, and determining the stability and repayment capacity of the cash flow to obtain a cash flow evaluation result;
S102, analyzing the funds inflow and outflow condition of the enterprise in detail by utilizing the integrated data set, monitoring the periodical change of the funds flow, such as seasonal and monthly fluctuation, and evaluating the stability of the funds flow by monitoring, particularly focusing on the maximum peak value and the minimum valley value of the funds flow at a specific time node (such as the end of a quarter or the end of a year), calculating and evaluating the repayment capability of the enterprise, namely the capability of the enterprise to pay on-demand debt in each time period, and predicting the financial performance and risk of the enterprise under different conditions by simulating the funds flow change under different economic environments.
S103, based on the fund liquidity assessment result, assessing the financial health state of the enterprise, calculating key financial ratios including the asset liability ratio and the yield by adopting industry standardized ratios, and obtaining a financial health index;
S103, comprehensively using a common financial ratio analysis method in the industry, such as asset liability ratio and profitability analysis, so as to measure the financial health state of the enterprise, combining the profitability of the enterprise (namely, the ratio of the total income to the total assets of the enterprise) with the repayment capacity of the enterprise (namely, the ratio of the total assets to the total liabilities of the enterprise) to form a comprehensive financial health index, wherein the index can clearly show the financial condition of the enterprise in the same industry, and help the enterprise understand the competitive position and the long-term development potential of the enterprise in the market.
Referring to fig. 3, based on the financial health index, the credit risk factor is extracted, the relevance of the lending behavior and the market change is quantitatively analyzed, the probability of the enterprise breach is calculated by the risk factor, and the step of obtaining the breach probability score is specifically as follows:
s201, identifying risk factors associated with enterprise default probabilities based on financial health indexes, including debt levels and income fluctuations, collecting factors and constructing a risk analysis framework to generate a risk analysis infrastructure;
S201, identifying key risk factors directly related to the default behavior of the enterprise on the basis of the financial health index, wherein the debt level and income fluctuation are particularly concerned, because the repayment capacity and the operation stability of the enterprise are directly influenced, collecting related data such as historical debt records, seasonal and periodical fluctuation data of income, and combining the result of the financial health index to construct a multi-dimensional risk analysis framework comprising indexes such as debt ratio, cash flow fluctuation and the like, and systematically analyzing and predicting the financial risk of the enterprise by the structure.
S202, analyzing the relevance of lending behavior and market dynamics based on a risk analysis basic structure, and extracting market sensitivity indexes by identifying the influence of market condition change on the lending behavior to obtain a market influence assessment result;
S202, deeply researching interaction between market dynamics and enterprise lending behaviors through a risk analysis infrastructure, focusing on direct influences of factors such as market economic conditions, such as interest rate change, market demand fluctuation, macroscopic economic policy adjustment and the like, on the enterprise lending behaviors, analyzing how the factors influence financial conditions and lending demands of enterprises, extracting indexes related to market sensitivity, such as influences of market fluctuation on enterprise cash flow and debt compensation ability, and in the process, revealing specific influences of market condition change on the enterprise lending behaviors through data analysis to form a detailed report of market influence evaluation.
S203, integrating risk factors and market dynamic data by adopting a weight distribution method based on a market influence evaluation result, and calculating the default probability of an enterprise under the current market condition to obtain a default probability score;
S203, comprehensively utilizing a market influence assessment result and a risk factor, measuring the contribution degree of the factors to the enterprise breach probability by setting the weight of each factor, and distributing the sensitivity and the actual influence of the factors to be considered by weight, wherein the sensitivity and the actual influence of the factors are different, such as the debt level and the influence of market fluctuation.
Referring to fig. 4, based on the breach probability score, by using a long-short term memory network and a weighted mahalanobis distance algorithm, through the financial stability and growth potential of an enterprise, a loan product with the optimal matching degree is identified, and the corresponding relation between the product condition and the enterprise requirement is analyzed, so that the step of obtaining the product matching index information is specifically as follows:
s301, based on the default probability score, screening financial indexes including profit margin and cash flow rate by adopting a long-short-period memory network, comparing the financial indexes with the characteristics of the existing loan products, identifying the matching degree of the loan products and the financial conditions of enterprises, and generating product matching efficiency analysis information;
S301, deeply analyzing financial data of enterprises, particularly profit margin and cash flow, by utilizing a long-short-period memory network, wherein indexes reflect the profitability and the fund liquidity of the enterprises, comparing the financial indexes with the characteristics of existing loan products on the market, such as fund limit, risk level and target client group of the products, identifying which financial requirements of the loan products are most consistent with the actual conditions of the enterprises through the comparison, further evaluating the matching degree of each product and the financial state of the enterprises, and generating a comprehensive analysis report about the matching efficiency of the loan products through collecting and analyzing a large number of data points in the process.
S302, based on product matching efficiency analysis information, comparing loan conditions of each product, including repayment deadlines and interest rates, with the fund requirements and compensation capabilities of enterprises by using a weighted Markov distance algorithm, and evaluating the matching effectiveness to obtain a loan adaptation degree checking result;
S302, comparing specific conditions of different loan products, such as repayment deadlines, interest rates and the like, carrying out matching analysis on the specific conditions with the fund requirements and the compensation capabilities of enterprises, finding out the closest loan options by calculating the distances between the characteristics of the different loan products and the financial requirements of the enterprises, ensuring that the recommended loan products are matched with the financial indexes, and matching the repayment capabilities and the fund arrangement of the enterprises on the repayment conditions, and finally generating an exhaustive loan adaptation evaluation result to help the enterprises to determine which loan conditions are most consistent with the actual conditions.
A weighted mahalanobis distance algorithm, according to the formula:
a weighted Marshall distance D WM between the loan product and the enterprise financial demand is calculated, where x is the target enterprise financial demand vector, y is the loan product feature vector, S is the covariance matrix of the overall enterprise financial demand, and W is the weight matrix.
The execution process comprises the following steps:
Firstly, a covariance matrix S and an inverse matrix S -1 of the total enterprise financial demand are calculated, then, importance of each loan condition and the enterprise demand is determined, for this purpose, a weight matrix W of each feature is obtained through historical data analysis, next, a difference value (x-y) and a transpose (x-y) T thereof in a target enterprise financial demand vector x and a feature vector y of a loan product are calculated, and again, the difference value transpose (x-y) T of the target enterprise financial demand vector x and the feature vector y of the loan product, the weight matrix W, the inverse matrix S -1 of the covariance matrix, the target enterprise financial demand vector x and the feature vector y of the loan product are multiplied, and finally, a square root is taken to obtain a weighted mahalanobis distance, and the distance is used for determining the loan product closest to the enterprise demand.
S303, integrating financial stability scores and growth potential evaluations of enterprises based on loan fitness verification results, listing fitness and potential risks of each recommended product, and generating product matching index information;
and S303, integrating the financial stability score and growth potential evaluation of the enterprise, wherein the score reflects the long-term development prospect and short-term financial health state of the enterprise, and based on the information, the matching degree and the facing potential risk of each recommended loan product are detailed, and the matching index information is used for detailed display of the advantages and disadvantages of each loan product, so that the enterprise can make an intelligent loan decision on the basis of comprehensively knowing various options, and the process ensures that the loan product selected by the enterprise can best meet the financial requirements of the enterprise, and considers the long-term growth potential and the risks brought by market change of the loan product.
Referring to fig. 5, based on product matching index information, a decision tree algorithm and a gating circulation unit are adopted to perform correlation analysis of credit levels and default risks, risk parameters are quantified for each credit level, enterprise risk levels are divided, and the steps of obtaining risk level classification results are specifically as follows:
S401, based on product matching index information, adopting a decision tree algorithm to collect enterprise historical default data of a plurality of credit grades, analyzing a risk mode of the data, identifying risk factors sensitive to the credit grades, and generating a credit risk factor set;
And S401, firstly, collecting historical default data of enterprises with different credit grades, including the frequency of default events, the amount of money involved and the recovery condition after default, analyzing the data by utilizing a decision tree algorithm to identify and verify the credit risk factors with the most influence, such as liability level, cash flow deficiency or profitability reduction, and forming a detailed credit risk factor set by constructing decision paths under different conditions by the decision tree to reveal the relation between the credit grades and the risk factors.
S402, based on a credit risk factor set, applying a gating circulation unit, setting a risk threshold for each level by comparing and analyzing risk exposure under a plurality of credit levels, and evaluating potential credit variation according to the threshold to obtain a credit risk quantification model;
And S402, processing time sequence data by using a gate control loop unit (GRU) algorithm, particularly, monitoring the change of risk factors at a plurality of time points, effectively capturing the dynamic change of the risk factors in different credit levels by the GRU through a memory and updating gate mechanism of the GRU, enabling analysis to accurately reflect the real credit state of an enterprise in an economic cycle, setting a specific risk threshold value for each credit level based on the analysis, predicting the future potential credit change, and finally integrating the threshold value and analysis result into a credit risk quantification model which can be used for monitoring and predicting the change of the credit state of the enterprise in real time.
S403, classifying risk grades of enterprises based on the credit risk quantification model, wherein the grades comprise low risk, medium risk and high risk, and designating matched loan conditions and interest rates for each grade to generate a risk grade classification result;
S403, classifying the enterprises into three grades of low risk, medium risk or high risk according to the output of the credit risk quantification model, wherein each risk grade corresponds to different loan conditions and interest rates, the conditions and interest rates are carefully designed according to the risk bearing capacity and repayment potential of the enterprises, the low-risk enterprises obtain more preferential interest rates and more flexible repayment conditions, the high-risk enterprises face higher interest rates and more strict loan conditions, and the risk grade classification result not only helps a loan organization reasonably configure resources.
Referring to fig. 6, based on the risk level classification result, the present loan structure is evaluated, and through industry comparison analysis, potential financing gaps and optimization points are identified, and the insufficient and potential improvement points in the loan structure are quantified, and the steps of generating the loan structure optimization record are specifically as follows:
s501, based on a risk level classification result, comparing consistency of enterprise multi-type loan conditions and risk levels, checking coverage and limitation conditions of loan products, identifying areas in excessive risk concentration, and generating a loan structure preliminary analysis record;
S501, firstly analyzing the matching degree between various loan products and enterprise risk grades, and mainly checking whether the loan conditions are suitable for all levels of risk enterprises, particularly checking whether the high-risk loan products reasonably cover related enterprise groups, in addition, evaluating whether the distribution of the loan products in different risk grades is balanced, identifying risk concentration areas, such as that certain high-risk loans are concentrated in specific industries or enterprise scales, and generating a preliminary analysis record containing current loan structure evaluation and potential risk points through analysis.
S502, based on the preliminary analysis record of the loan structure, using industry average data as a reference, evaluating the deviation degree of the enterprise loan structure and industry average, and determining financing gaps and structure shortages to obtain financing gaps and an optimization scheme;
S502, comparing the loan structure of the enterprise with the industry average level, analyzing the differences in loan types, conditions and risk bearing capacity, particularly in loan product diversity, the loose degree of loan conditions and risk preventive measures, comparing and revealing main gaps and shortages of the enterprise in the financing structure, such as existing funds shortage or overrelying on a certain loan type, and preparing a targeted financing gap and structure optimization scheme according to the analysis result.
S503, based on the financing gap and the optimization scheme, setting optimization measures including loan condition adjustment, new loan product introduction or existing product condition modification, matching the requirements and risk bearing capacity of enterprises, and generating a loan structure optimization record;
And S503, implementing specific optimization measures according to the analysis result and the optimization scheme, wherein the specific optimization measures comprise adjusting the interest rate and repayment condition of the existing loan products, introducing new loan products meeting the current market demands and enterprise risk characteristics, or carrying out necessary modification on the existing products to better adapt to the enterprise demands of different risk grades, optimizing the loan structure of the enterprise through the measures, improving the financing efficiency and the risk management capability of the enterprise, and the loan structure optimization record generated in the process details each implemented measure and expected effect.
Referring to fig. 7, based on the loan structure optimization record, the steps of performing dynamic risk assessment, predicting the trend of variation of liabilities, updating the risk assessment model and matching the market variation by using the historical data of financial performance, and obtaining the dynamic monitoring result of loan risk are specifically as follows:
S601, based on loan structure optimization records, collecting historical financial data and current market dynamic information of enterprises, analyzing the influence of the data on the enterprise property liability ratio and cash flow, identifying the change trend of financial performance, and generating market and financial dynamic analysis information;
and S601, firstly, collecting historical financial data of enterprises, including a liability list, a cash flow list, a damage list and the like of the past years according to the optimized loan structure record, and simultaneously, considering the current market dynamics, such as factors of interest rate change, economic growth rate, industry competition condition and the like, evaluating specific influences on the liability ratio and cash flow of the enterprises by analyzing the data, identifying main change trend of the financial performance of the enterprises, integrating the information into a market and financial dynamic analysis report, and providing real-time data support for risk evaluation.
S602, based on market and financial dynamic analysis information, evaluating the accuracy of the asset liability prediction model, dynamically adjusting risk parameters according to the market, and optimizing the response capability of the model to obtain a risk evaluation model update record;
S602, evaluating an existing asset liability prediction model by utilizing collected market and financial dynamic analysis information, wherein the evaluation comprises the steps of verifying the prediction accuracy of the model and the response speed to new data, and adjusting key risk parameters in the model according to the current and predicted market conditions, such as the sensitivity to economic periodical changes, the influence of interest rate changes and the like, so as to improve the response capability of the model to market changes, wherein the update records generated in the process illustrate the specific content and expected effect of model adjustment.
S603, updating records based on a risk assessment model, performing risk monitoring, reflecting changes of market and financial conditions in real time, optimizing real-time performance of risk early warning and management measures, and generating a loan risk dynamic monitoring result;
And S603, performing continuous risk monitoring by using the updated risk assessment model, wherein the continuous risk monitoring comprises real-time tracking of market changes and financial conditions of enterprises, such as fluctuation of property liability ratio, fluctuation of cash flow and the like, and automatically generating risk early warning according to a set threshold value and an alarm mechanism to help the enterprises to timely identify and cope with potential financial risks, and the final output of the process is a loan risk dynamic monitoring result, so that a loan structure can be effectively optimized and potential financial pressure can be relieved.
Referring to fig. 8, an artificial intelligence based loan risk assessment system, the system comprising:
the data integration module integrates historical loan records, financial reports, bank flowing water and market behavior data of enterprises, analyzes the fund flowing and debt repaying capacity, evaluates the financial conditions of the enterprises and generates financial health indexes;
The credit risk assessment module extracts credit risk factors based on the financial health index, quantitatively analyzes the relevance of lending behaviors and market changes, calculates the enterprise default probability, and obtains a default probability score;
The product matching analysis module evaluates the financial stability and growth potential of enterprises based on the default probability score, identifies and matches the optimal loan products, analyzes the corresponding relation between the product conditions and the enterprise demands, and acquires product matching index information;
the risk level classification module analyzes the relevance of the credit level and the default risk by using a decision tree algorithm and a gating circulation unit according to the product matching index information, quantifies the risk parameter of each credit level, and classifies the enterprise risk level to obtain a risk level classification result;
The loan structure optimization module evaluates the current loan structure based on the risk level classification result, identifies potential financing gaps and optimization points in industry comparison, quantifies defects in the loan structure, updates the risk assessment model and matches market changes, and generates a loan structure optimization record.
The present invention is not limited to the above embodiments, and any equivalent embodiments which can be changed or modified by the technical disclosure described above can be applied to other fields, but any simple modification, equivalent changes and modification made to the above embodiments according to the technical matter of the present invention will still fall within the scope of the technical disclosure.
Claims (8)
1. An artificial intelligence-based loan risk assessment method is characterized by comprising the following steps:
Collecting historical loan records, financial reports, bank flowing water and market behavior data of enterprises, integrating the collected data, analyzing the flowability and repaying capacity of funds, evaluating the financial conditions of the enterprises and generating financial health indexes;
based on the financial health index, extracting credit risk factors, quantitatively analyzing the relevance of lending behaviors and market changes, and calculating the default probability of enterprises through the risk factors to obtain default probability scores;
based on the default probability score, identifying loan products with optimal matching degree by utilizing a long-short-term memory network and a weighted mahalanobis distance algorithm, and analyzing the corresponding relation between product conditions and enterprise requirements to obtain product matching index information;
Based on the product matching index information, adopting a decision tree algorithm and a gating circulation unit to perform correlation analysis of credit levels and default risks, quantifying risk parameters for each credit level, and obtaining a risk level classification result;
based on the risk level classification result, evaluating the current loan structure, and identifying potential financing gaps and optimization points through industry comparison analysis to generate a loan structure optimization record;
Based on the loan structure optimization record, carrying out dynamic risk assessment by utilizing historical data of financial performance, and predicting the change trend of the liability to obtain a loan risk dynamic monitoring result;
based on the default probability score, a loan product with optimal matching degree is identified by utilizing a long-short-term memory network and a weighted mahalanobis distance algorithm, and the corresponding relation between product conditions and enterprise requirements is analyzed, so that the step of obtaining product matching index information is specifically as follows:
based on the default probability score, screening financial indexes including profit margin and cash flow rate by adopting a long-short-period memory network, comparing the financial indexes with the characteristics of the existing loan products, identifying the matching degree of the loan products and the financial conditions of enterprises, and generating product matching efficiency analysis information;
based on the product matching efficiency analysis information, comparing loan conditions of each product, including repayment deadlines and interest rates, with the fund requirements and compensation capabilities of enterprises by using a weighted mahalanobis distance algorithm, and evaluating the matching effectiveness to obtain a loan adaptation degree checking result;
based on the loan fitness verification result, integrating financial stability scores and growth potential evaluations of enterprises, listing fitness and potential risks of each recommended product, and generating product matching index information;
based on the product matching index information, adopting a decision tree algorithm and a gating circulation unit to perform correlation analysis of credit levels and default risks, quantifying risk parameters for each credit level, and obtaining a risk level classification result specifically comprises the following steps:
Based on the product matching index information, adopting a decision tree algorithm to collect enterprise historical default data with a plurality of credit grades, analyzing a risk mode of the data, identifying risk factors sensitive to the credit grades, and generating a credit risk factor set;
Based on the credit risk factor set, a gating circulation unit is applied, risk exposure under a plurality of credit levels is compared and analyzed, a risk threshold is set for each level, potential credit variation is evaluated according to the threshold, and a credit risk quantification model is obtained;
and carrying out risk classification on enterprises based on the credit risk quantification model, wherein the classes comprise low risk, medium risk and high risk, and designating matched loan conditions and interest rates for each hierarchy to generate a risk class classification result.
2. The artificial intelligence based loan risk assessment method of claim 1, wherein the financial health index comprises a financial structural index, a repayment capability index, a performance capability index, a profitability index, a development capability index, the offensiveness score comprises a credit score, financial condition change information, and a market behavior analysis result, the product matching index information comprises a loan category, a guarantee measure, a loan deadline, and a interest rate option, the risk level classification result comprises an offending risk rating, an expected loss rate, and a credit recovery potential, the loan structural optimization record comprises an adjusted loan proportion, a risk dispersion policy, and a debt reorganization plan, and the loan risk dynamic monitoring result comprises an adjusted offending probability prediction, a market fluctuation response, and a risk mitigation measure update result.
3. The method for evaluating loan risk based on artificial intelligence according to claim 1, wherein the steps of collecting historical loan records, financial statements, banking running water and market behavior data of the enterprise, integrating the collected data, analyzing the mobility of funds and the ability of repayment, and evaluating the financial condition of the enterprise, and generating the financial health index are specifically as follows:
collecting enterprise historical loan records, financial reports, bank flowing water and market behavior data, and cleaning data, including consistency checking, removing repeated items, correcting error formats and filling missing values, to generate an integrated data set;
Based on the integrated data set, analyzing cash inflow and outflow modes of enterprises, evaluating the fluctuation of the cash flow of month and year, and determining the stability and repayment capacity of the cash flow to obtain a cash flow evaluation result;
based on the fund liquidity assessment result, assessing the financial health state of the enterprise, calculating key financial ratios including the asset liability ratio and the yield by adopting industry standardized ratios, and obtaining a financial health index.
4. The method for evaluating loan risk based on artificial intelligence according to claim 1, wherein the steps of extracting credit risk factors based on the financial health index, quantitatively analyzing the association of loan behaviors with market changes, and calculating the probability of violations of enterprises by the risk factors to obtain the probability scores of violations are specifically as follows:
Identifying risk factors associated with enterprise breach probabilities, including debt levels and funding volatility, based on the financial health index, collecting the factors and constructing a risk analysis framework, generating a risk analysis infrastructure;
Based on the risk analysis basic structure, the relevance of the lending behavior and market dynamics is analyzed, and the market sensitivity index is extracted by identifying the influence of the market condition change on the lending behavior, so that a market influence assessment result is obtained;
And based on the market influence evaluation result, integrating the risk factors and the market dynamic data by adopting a weight distribution method, and calculating the default probability of the enterprise under the current market condition to obtain a default probability score.
5. The method of claim 1, wherein the weighted mahalanobis distance algorithm is according to the formula:
Calculating a weighted mahalanobis distance between a loan product and an enterprise financial demand Wherein, the method comprises the steps of, wherein,Is a vector of financial needs of the target enterprise,In order to loan the product feature vector,Is the covariance matrix of the financial requirements of the overall enterprise,Is a weight matrix.
6. The method of claim 1, wherein the step of evaluating the current loan structure based on the risk level classification result, identifying potential financing gaps and optimization points by industry comparison analysis, and generating a loan structure optimization record is specifically:
Based on the risk level classification result, comparing the consistency of the enterprise multi-class loan conditions and the risk level, checking the coverage range and the limiting conditions of the loan products, identifying the area in which excessive risks are concentrated, and generating a loan structure preliminary analysis record;
Based on the preliminary analysis record of the loan structure, using industry average data as a benchmark, evaluating the deviation degree of the enterprise loan structure and industry average, and determining financing gaps and structure shortages to obtain financing gaps and an optimization scheme;
Based on the financing gap and the optimization scheme, optimization measures are formulated, including loan condition adjustment, new loan product introduction or existing product condition modification, and requirements and risk bearing capacity of enterprises are matched to generate a loan structure optimization record.
7. The method for evaluating loan risk based on artificial intelligence according to claim 1, wherein the step of performing dynamic risk evaluation based on the loan structure optimization record by using historical data of financial performance to predict the trend of variation of liability and obtaining the dynamic monitoring result of loan risk comprises the following steps:
based on the loan structure optimization record, collecting historical financial data and current market dynamic information of enterprises, analyzing the influence of the data on the enterprise property liability ratio and cash flow, identifying the change trend of financial performance, and generating market and financial dynamic analysis information;
based on the market and financial dynamic analysis information, evaluating the accuracy of the asset liability prediction model, dynamically adjusting risk parameters according to the market, and optimizing the response capability of the model to obtain a risk evaluation model update record;
And updating records based on the risk assessment model, performing risk monitoring, reflecting the changes of market and financial conditions in real time, optimizing the real-time performance of risk early warning and management measures, and generating a loan risk dynamic monitoring result.
8. An artificial intelligence based loan risk assessment system, wherein the artificial intelligence based loan risk assessment method according to any one of claims 1-7 is performed, the system comprising:
the data integration module integrates historical loan records, financial reports, bank flowing water and market behavior data of enterprises, analyzes the fund flowing and debt repaying capacity, evaluates the financial conditions of the enterprises and generates financial health indexes;
The credit risk assessment module extracts credit risk factors based on the financial health index, quantitatively analyzes the relevance of lending behaviors and market changes, calculates the enterprise default probability, and obtains a default probability score;
The product matching analysis module evaluates the financial stability and growth potential of enterprises based on the default probability score, identifies and matches the optimal loan products, analyzes the corresponding relation between the product conditions and the enterprise demands, and acquires product matching index information;
The risk level classification module analyzes the relevance of the credit level and the default risk by using a decision tree algorithm and a gating circulation unit according to the product matching index information, quantifies the risk parameter of each credit level, and classifies the enterprise risk level to obtain a risk level classification result;
And the loan structure optimization module evaluates the current loan structure based on the risk level classification result, identifies potential financing gaps and optimization points in industry comparison, quantifies defects in the loan structure, updates a risk assessment model and matches market changes, and generates a loan structure optimization record.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410779852.9A CN118799058B (en) | 2024-06-17 | 2024-06-17 | A loan risk assessment method and system based on artificial intelligence |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410779852.9A CN118799058B (en) | 2024-06-17 | 2024-06-17 | A loan risk assessment method and system based on artificial intelligence |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN118799058A CN118799058A (en) | 2024-10-18 |
| CN118799058B true CN118799058B (en) | 2025-04-15 |
Family
ID=93030972
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202410779852.9A Active CN118799058B (en) | 2024-06-17 | 2024-06-17 | A loan risk assessment method and system based on artificial intelligence |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN118799058B (en) |
Families Citing this family (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN119444455A (en) * | 2024-10-25 | 2025-02-14 | 北京数聚智连科技股份有限公司 | A financial data analysis and processing method and system |
| CN119831721B (en) * | 2024-11-27 | 2025-10-28 | 中国工商银行股份有限公司 | Credit risk control method and device based on multi-source data fusion and reinforcement learning |
| CN119250549B (en) * | 2024-12-09 | 2025-04-25 | 上海银行股份有限公司 | Enterprise financial health assessment method based on big data |
| CN119313449B (en) * | 2024-12-13 | 2025-03-28 | 江苏金农股份有限公司 | Multi-source information analysis driven intelligent repayment strategy generation system and method |
| CN119762218A (en) * | 2024-12-27 | 2025-04-04 | 中国工商银行股份有限公司 | Risk prediction method, device, storage medium and electronic device for loan business |
| CN119692784A (en) * | 2025-02-21 | 2025-03-25 | 长沙淘顺信息科技有限公司 | Intelligent business risk control strategy management system |
| CN119850331A (en) * | 2025-03-19 | 2025-04-18 | 杭州坤威信息科技有限公司 | Bad asset assessment system and method |
| CN120197949A (en) * | 2025-04-01 | 2025-06-24 | 广东中盈盛达数字科技有限公司 | Enterprise risk assessment method and system based on large language model |
| CN120494963A (en) * | 2025-05-09 | 2025-08-15 | 江苏幺贰凌玖科技服务有限公司 | Bank credit risk management method based on big data analysis |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111654866A (en) * | 2020-05-29 | 2020-09-11 | 北京合力思腾科技股份有限公司 | Method, device and computer storage medium for preventing mobile communication from fraud |
| CN112966768A (en) * | 2021-03-19 | 2021-06-15 | 廊坊银行股份有限公司 | User data classification method, device, equipment and medium |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN118096360A (en) * | 2024-04-28 | 2024-05-28 | 深圳市拓保软件有限公司 | Intelligent banking risk early warning test method based on hong Mongolian system |
| CN118134633A (en) * | 2024-05-07 | 2024-06-04 | 深圳市拓保软件有限公司 | Bank credit asset risk analysis method and system based on AI technology |
-
2024
- 2024-06-17 CN CN202410779852.9A patent/CN118799058B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111654866A (en) * | 2020-05-29 | 2020-09-11 | 北京合力思腾科技股份有限公司 | Method, device and computer storage medium for preventing mobile communication from fraud |
| CN112966768A (en) * | 2021-03-19 | 2021-06-15 | 廊坊银行股份有限公司 | User data classification method, device, equipment and medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN118799058A (en) | 2024-10-18 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN118799058B (en) | A loan risk assessment method and system based on artificial intelligence | |
| US8577791B2 (en) | System and computer program for modeling and pricing loan products | |
| Huang et al. | Contractor financial prequalification using simulation method based on cash flow model | |
| CN118941381B (en) | A post-loan risk warning and disposal method for automobile finance | |
| Bandyopadhyay | Managing portfolio credit risk in banks: An Indian perspective | |
| Kolodiziev et al. | Use of causal analysis to improve the monitoring of the banking system stability | |
| CN118071513B (en) | Personalized bank asset management method and system based on artificial intelligence driving | |
| Kedarya et al. | Calculating strategic risk in financial institutions | |
| CN112766814A (en) | Training method, device and equipment for credit risk pressure test model | |
| CN116596674A (en) | External trade risk assessment method based on big data analysis | |
| Tikhonov et al. | The Relationship Between the Financial Performance of Banks and the Quality of Credit Scoring Models | |
| CN120317977A (en) | A dynamic financing risk management method and system based on data model | |
| Boguslauskas et al. | The selection of financial ratios as independent variables for credit risk assessment | |
| CN119809792A (en) | A credit limit dynamic management system and method based on AI big model | |
| Vojtková et al. | Intraday liquidity modelling using statistical methods | |
| Xu et al. | High-frequency Growth-at-Risk of China: the Role of Macro-financial Environment | |
| CN115619525A (en) | A method and system for enterprise post-loan early warning based on electricity consumption data | |
| Gök et al. | Dating currency crisis and assessing the determinants based on meta fuzzy index functions | |
| Linden | Synthesis of research studies examining prediction of bankruptcy | |
| Oppusunggu et al. | Risk Profile Assessment ofRisk profile assessment of core capital adequacy: capital adequacy tier-book strategy Core Capital Adequacy: Capital Adequacy Tier-Book Strategy | |
| Kanyambu et al. | Modelling credit risk using system dynamics: The case of licensed credit reference bureaus in Kenya | |
| Cusano et al. | Learning from revisions: an algorithm to detect errors in banks’ balance sheet statistical reporting | |
| Dabi | Modelling Corporate Probability of Default–A Possible Supervisory Benchmark Model | |
| Parnes | Approximating default probabilities with soft information | |
| ŢIŢAN et al. | Conceptual and statistical issues regarding the probability of default and modeling default risk |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |