CN113822490B - Asset collection method and device based on artificial intelligence and electronic equipment - Google Patents
Asset collection method and device based on artificial intelligence and electronic equipment Download PDFInfo
- Publication number
- CN113822490B CN113822490B CN202111155958.4A CN202111155958A CN113822490B CN 113822490 B CN113822490 B CN 113822490B CN 202111155958 A CN202111155958 A CN 202111155958A CN 113822490 B CN113822490 B CN 113822490B
- Authority
- CN
- China
- Prior art keywords
- client
- algorithm model
- feature
- features
- candidate
- 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
- 238000000034 method Methods 0.000 title claims abstract description 72
- 238000013473 artificial intelligence Methods 0.000 title claims abstract description 41
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 118
- 238000007477 logistic regression Methods 0.000 claims abstract description 92
- 238000003066 decision tree Methods 0.000 claims abstract description 41
- 238000009472 formulation Methods 0.000 claims abstract description 10
- 239000000203 mixture Substances 0.000 claims abstract description 10
- 238000012795 verification Methods 0.000 claims description 56
- 238000012545 processing Methods 0.000 claims description 31
- 238000012549 training Methods 0.000 claims description 29
- 238000011156 evaluation Methods 0.000 claims description 24
- 238000003860 storage Methods 0.000 claims description 23
- 230000000694 effects Effects 0.000 claims description 13
- 238000012216 screening Methods 0.000 claims description 11
- 238000007781 pre-processing Methods 0.000 claims description 10
- 238000004364 calculation method Methods 0.000 claims description 9
- 238000004140 cleaning Methods 0.000 claims description 6
- 230000001737 promoting effect Effects 0.000 claims 1
- 238000005516 engineering process Methods 0.000 abstract description 14
- 238000004590 computer program Methods 0.000 description 12
- 238000010586 diagram Methods 0.000 description 11
- 230000008569 process Effects 0.000 description 11
- 230000006870 function Effects 0.000 description 8
- 238000004891 communication Methods 0.000 description 6
- 238000007726 management method Methods 0.000 description 5
- 238000001914 filtration Methods 0.000 description 4
- 230000003287 optical effect Effects 0.000 description 4
- 230000002159 abnormal effect Effects 0.000 description 3
- 238000004458 analytical method Methods 0.000 description 3
- 230000001419 dependent effect Effects 0.000 description 3
- 238000013459 approach Methods 0.000 description 2
- 230000006399 behavior Effects 0.000 description 2
- 239000004973 liquid crystal related substance Substances 0.000 description 2
- 238000010801 machine learning Methods 0.000 description 2
- 230000000644 propagated effect Effects 0.000 description 2
- 239000004065 semiconductor Substances 0.000 description 2
- 206010043417 Therapeutic response unexpected Diseases 0.000 description 1
- 230000006978 adaptation Effects 0.000 description 1
- 150000001875 compounds Chemical class 0.000 description 1
- 238000007405 data analysis Methods 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000003306 harvesting Methods 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 238000011835 investigation Methods 0.000 description 1
- 238000012886 linear function Methods 0.000 description 1
- 238000012417 linear regression Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000003058 natural language processing Methods 0.000 description 1
- 239000013307 optical fiber Substances 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000012546 transfer Methods 0.000 description 1
- 230000001960 triggered effect Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
- G06F17/10—Complex mathematical operations
- G06F17/18—Complex mathematical operations for evaluating statistical data, e.g. average values, frequency distributions, probability functions, regression analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/243—Classification techniques relating to the number of classes
- G06F18/24323—Tree-organised classifiers
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q40/00—Finance; Insurance; Tax strategies; Processing of corporate or income taxes
- G06Q40/03—Credit; Loans; Processing thereof
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Business, Economics & Management (AREA)
- Economics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Strategic Management (AREA)
- Life Sciences & Earth Sciences (AREA)
- Evolutionary Biology (AREA)
- General Engineering & Computer Science (AREA)
- Operations Research (AREA)
- Finance (AREA)
- Evolutionary Computation (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Marketing (AREA)
- Artificial Intelligence (AREA)
- Computational Mathematics (AREA)
- Human Resources & Organizations (AREA)
- General Business, Economics & Management (AREA)
- Mathematical Analysis (AREA)
- Mathematical Optimization (AREA)
- Accounting & Taxation (AREA)
- Pure & Applied Mathematics (AREA)
- Development Economics (AREA)
- Mathematical Physics (AREA)
- Databases & Information Systems (AREA)
- Software Systems (AREA)
- Algebra (AREA)
- Probability & Statistics with Applications (AREA)
- Game Theory and Decision Science (AREA)
- Technology Law (AREA)
- Entrepreneurship & Innovation (AREA)
- Quality & Reliability (AREA)
- Tourism & Hospitality (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The application relates to an artificial intelligence technology, and particularly discloses an asset collection method and device based on artificial intelligence, and electronic equipment, wherein the method comprises the following steps: acquiring a customer feature library associated with the asset to be cleared; inputting the client feature library into a preset decision tree algorithm model, so that the decision tree algorithm model obtains importance values of client features contained in the client feature library, and selecting a specified number of client features as candidate client features according to the sequence of the importance values from large to small, wherein the importance values are used for representing the importance degree of the client features relative to asset collection strategy formulation; inputting the obtained candidate client features into a preset logistic regression algorithm model to obtain a prediction result output by the logistic regression algorithm model; and determining the client type with the maximum probability value from the prediction result, and selecting an asset collection strategy corresponding to the client type with the maximum probability value. By adopting the technical scheme of the embodiment of the application, a more reliable asset collection strategy can be obtained.
Description
Technical Field
The application relates to the technical field of artificial intelligence, in particular to an asset collection method and device based on artificial intelligence.
Background
The processing of the financial property by the bank includes clearing the property of the customer, for example, the customer obtains the loan from the bank and returns due to the self-operation income, but if the operation income of the customer is insufficient to repay the loan of the bank, the loan becomes an abnormal loan, and the bank needs to take measures to recover the loan.
At present, a bank mainly relies on professionals responsible for clearing and collecting in the bank to perform subjective evaluation, clients with poor evaluation are entrusted to a third party service company to perform clearing and collecting, and clients with good evaluation are cleared and collected in a mode of collecting in the bank, so that the efficiency of the manual clearing and collecting mode is low, and workers usually subjectively select to entrust most clients to the third party service company, so that clearing and collecting cost of the bank is very high.
Disclosure of Invention
To solve the above technical problems, embodiments of the present application provide an asset collection method and device based on artificial intelligence, an electronic device, and a computer readable storage medium, so as to improve the efficiency and reliability of collection decision.
Other features and advantages of the application will be apparent from the following detailed description, or may be learned by the practice of the application.
According to an aspect of an embodiment of the present application, there is provided an asset collection method based on artificial intelligence, including: acquiring a customer feature library associated with the asset to be cleared; inputting the client feature library into a preset decision tree algorithm model, so that the decision tree algorithm model obtains importance values of client features contained in the client feature library, and selecting a specified number of client features as candidate client features according to the sequence of the importance values from large to small, wherein the importance values are used for representing importance degrees of the client features relative to asset collection strategy formulation; inputting the obtained candidate client features into a preset logistic regression algorithm model to obtain a prediction result output by the logistic regression algorithm model, wherein the prediction result comprises a first probability corresponding to a first client type and a second probability corresponding to a second client type; and determining the client type with the maximum probability value from the prediction result, and selecting an asset collection strategy corresponding to the client type with the maximum probability value as a target strategy for asset collection for the client.
In another exemplary embodiment, after inputting the client feature library into a preset decision tree algorithm model, so that the decision tree algorithm model obtains importance values of client features contained in the client feature library, and selecting a specified number of client features from a large-to-small order according to the importance values as candidate client features, the method further includes: calculating information values corresponding to the candidate client features, and displaying the obtained calculation results, wherein the information value user represents the contribution degree of the candidate client features relative to the asset collection strategy; and screening the candidate client features from the candidate client features according to the user control information input in real time so as to input the screened candidate client features into the logistic regression algorithm model, wherein the user control information is used for representing client feature selection operations made by expert users based on self-cleaning experience.
In another exemplary embodiment, the calculating the information value corresponding to each candidate client feature includes: grouping processing is carried out on each candidate client feature respectively so as to obtain feature groups corresponding to each candidate client feature respectively; calculating information values corresponding to each feature group based on feature information distribution in each feature group; and calculating the information values corresponding to the candidate features according to the information values corresponding to all the feature groups.
In another exemplary embodiment, before inputting the client feature library into a preset decision tree algorithm model, so that the decision tree algorithm model obtains importance values of client features contained in the client feature library, and selecting a specified number of client features from a large-to-small order according to the importance values as candidate client features, the method further includes: and carrying out data preprocessing on the client features contained in the client feature library so as to input the client feature library obtained after the processing into the decision tree algorithm model, wherein the data preprocessing comprises at least one of data processing and data screening.
In another exemplary embodiment, the method further comprises, prior to inputting the resulting candidate customer features into the pre-set logistic regression algorithm model: selecting asset collection data handed over in a first time period from historical asset collection data as a training sample set, and selecting asset collection data handed over in a second time period as a verification sample set, wherein the first time period is earlier than the second time period; training the logistic regression algorithm model to be trained according to the training sample set, and verifying the trained logistic regression algorithm model through the verification sample set to obtain a verification result; and if the verification result indicates that the trained logistic regression algorithm model meets the requirement, inputting the candidate client features into the trained logistic regression algorithm model.
In another exemplary embodiment, the method further comprises: and if the verification result indicates that the trained logistic regression algorithm model does not meet the requirements, repeating the steps of training the logistic regression algorithm model to be trained according to the training sample set, verifying the trained logistic regression algorithm model through the verification sample set to obtain the verification result until the obtained verification result indicates that the trained logistic regression algorithm model meets the requirements.
In another exemplary embodiment, the verifying the trained logistic regression algorithm model through the verification sample set to obtain a verification result includes: obtaining a prediction result output by the trained logistic regression algorithm model aiming at each verification sample in the verification sample set; and calculating a model effect evaluation value according to the obtained prediction result, and if the model effect evaluation value is larger than a preset threshold value, generating a verification result for indicating that the trained logistic regression algorithm model meets the requirement.
According to an aspect of an embodiment of the present application, there is provided an asset collection device based on artificial intelligence, including: the feature acquisition module is configured to acquire a customer feature library associated with the asset to be cleared; the feature selection module is configured to input the customer feature library into a preset decision tree algorithm model, so that the decision tree algorithm model obtains importance values of the customer features contained in the customer feature library, and a specified number of customer features are selected as candidate customer features according to the sequence of the importance values from large to small, wherein the importance values are used for representing importance degrees of the customer features formulated relative to an asset collection strategy; the probability prediction module is configured to input the obtained candidate client features into a preset logistic regression algorithm model to obtain a prediction result output by the logistic regression algorithm model, wherein the prediction result comprises a first probability corresponding to a first client type and a second probability corresponding to a second client type; and the strategy selection module is configured to determine the client type with the maximum probability value from the prediction result, and select an asset collection strategy corresponding to the client type with the maximum probability value as a target strategy for asset collection for the client.
According to an aspect of an embodiment of the present application, there is provided an electronic apparatus including: one or more processors; and storage means for storing one or more programs which, when executed by the one or more processors, cause the electronic device to implement an artificial intelligence based asset collection method as described above.
According to an aspect of an embodiment of the present application, there is provided a computer-readable storage medium having stored thereon computer-readable instructions which, when executed by a processor of a computer, cause the computer to perform the artificial intelligence based asset harvesting method as described above.
According to an aspect of embodiments of the present application, there is provided a computer program product or computer program comprising computer instructions stored in a computer readable storage medium. The processor of the computer device reads the computer instructions from the computer-readable storage medium and executes the computer instructions to cause the computer device to perform the artificial intelligence based asset collection method provided in the various alternative embodiments described above.
In the technical scheme provided by the embodiment of the application, the intelligent formulation of the customer clearing strategy in the asset clearing stage is realized based on the decision tree and logistic regression architecture, and compared with the prior art that the customer clearing strategy is selected by subjective evaluation by professionals responsible for clearing in banks, the technical scheme provided by the application can avoid artificial subjective influence and rely on the processing of customer characteristics related to asset clearing, so that the obtained asset clearing strategy is more reliable and has higher efficiency.
It is to be understood that both the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the application as claimed.
Drawings
The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the application and together with the description, serve to explain the principles of the application. It is evident that the drawings in the following description are only some embodiments of the present application and that other drawings may be obtained from these drawings without inventive effort for a person of ordinary skill in the art. In the drawings:
FIG. 1 is a schematic illustration of an implementation environment in which the present application is directed;
FIG. 2 is a flow chart illustrating an artificial intelligence based asset collection method according to an exemplary embodiment of the application;
FIG. 3 is a flow chart illustrating an artificial intelligence based asset collection method according to another exemplary embodiment of the application;
FIG. 4 is a flow chart illustrating an artificial intelligence based asset collection method according to another exemplary embodiment of the application;
FIG. 5 is a flow chart of step S310 in the embodiment shown in FIG. 4 in an exemplary embodiment;
FIG. 6 is a flow chart illustrating an artificial intelligence based asset collection method according to another exemplary embodiment of the application;
FIG. 7 is a block diagram of an asset collection device based on artificial intelligence, as shown in an exemplary embodiment of the application;
Fig. 8 shows a schematic diagram of a computer system suitable for use in implementing an embodiment of the application.
Detailed Description
Reference will now be made in detail to exemplary embodiments, examples of which are illustrated in the accompanying drawings. When the following description refers to the accompanying drawings, the same numbers in different drawings refer to the same or similar elements, unless otherwise indicated. The implementations described in the following exemplary examples do not represent all implementations consistent with the application. Rather, they are merely examples of apparatus and methods consistent with aspects of the application as detailed in the accompanying claims.
The block diagrams depicted in the figures are merely functional entities and do not necessarily correspond to physically separate entities. That is, the functional entities may be implemented in software, or in one or more hardware modules or integrated circuits, or in different networks and/or processor devices and/or microcontroller devices.
The flow diagrams depicted in the figures are exemplary only, and do not necessarily include all of the elements and operations/steps, nor must they be performed in the order described. For example, some operations/steps may be decomposed, and some operations/steps may be combined or partially combined, so that the order of actual execution may be changed according to actual situations.
In the present application, the term "plurality" means two or more. "and/or" describes an association relationship of an association object, meaning that there may be three relationships, e.g., a and/or B may represent: a exists alone, A and B exist together, and B exists alone. The character "/" generally indicates that the context-dependent object is an "or" relationship.
First, artificial intelligence (ARTIFICIAL INTELLIGENCE, AI) is a theory, method, technique, and application system that simulates, extends, and extends human intelligence, senses the environment, obtains knowledge, and uses knowledge to obtain optimal results using a digital computer or a machine controlled by a digital computer. In other words, artificial intelligence is an integrated technology of computer science that attempts to understand the essence of intelligence and to produce a new intelligent machine that can react in a similar way to human intelligence. Artificial intelligence, i.e. research on design principles and implementation methods of various intelligent machines, enables the machines to have functions of sensing, reasoning and decision.
The artificial intelligence technology is a comprehensive subject, and relates to the technology with wide fields, namely the technology with a hardware level and the technology with a software level. Artificial intelligence infrastructure technologies generally include technologies such as sensors, dedicated artificial intelligence chips, cloud computing, distributed storage, big data processing technologies, operation/interaction systems, mechatronics, and the like. The artificial intelligence software technology mainly comprises a computer vision technology, a voice processing technology, a natural language processing technology, machine learning/deep learning and other directions.
The asset collection method and device based on artificial intelligence, the electronic device and the computer readable storage medium according to the embodiments of the present application mainly relate to machine learning technology included in artificial intelligence technology, and these embodiments will be described in detail below.
Referring first to fig. 1, fig. 1 is a schematic diagram of an implementation environment according to the present application. The implementation environment includes a terminal 10 and a server 20, and communication is performed between the terminal 10 and the server 20 through a wired or wireless network.
The server 20 is configured to perform background processing on a client feature library corresponding to a client to be subjected to asset collection, so as to obtain an asset collection policy suitable for the client, and transmit the obtained asset collection policy to the terminal 10 for display. The bank staff can collect the assets according to the strategy information displayed in the terminal 10, and compared with the prior art, which manually selects the asset collection strategy, the asset collection strategy selection mode provided by the implementation environment is more reliable and has higher efficiency.
It should be noted that, the terminal 10 in the implementation environment shown in fig. 1 may be any electronic device such as a smart phone, a tablet, a notebook computer, a computer, etc.; the server 20 may be an independent server, or may be a cloud server that provides cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, content delivery networks (Content Delivery Network, CDN), and basic cloud computing services such as big data and artificial intelligence platforms, which are not limited in this regard.
FIG. 2 is a flow chart illustrating an artificial intelligence based asset collection method according to an exemplary embodiment of the application. The method may be applied to the implementation environment shown in fig. 1 and is specifically performed by the server 20 in the embodiment environment shown in fig. 1. In other embodiments, the method may be performed by a device in other embodiments, and this embodiment is not limited in this regard.
As shown in fig. 2, in an exemplary embodiment, the asset collection method based on artificial intelligence may include steps S110 to S170, which are described in detail as follows:
step S110, a customer feature library associated with the asset to be cleared is acquired.
First, the customer feature library associated with the asset to be collected refers to a feature set corresponding to a customer to be collected. After the customer to be cleared is handed over to the management department of the banker's assets, a lot of customer-related data is obtained to construct a customer profile library from the data.
Illustratively, the customer characteristic library includes at least one of a customer attribute characteristic, a customer property characteristic, a customer liability characteristic, a customer transaction characteristic, and a customer collection characteristic. Wherein, the customer attribute characteristics comprise information such as age, sex, overdue days, delinquent amount and the like of the customer; customer property characteristics include, for example, information on the customer's cash, property, car credits, house credits, equity, property rights, etc.; the customer liability feature includes, for example, information such as a customer's loan, debit card, civil loan, credit investigation data, etc.; customer transaction characteristics include, for example, repayment behavior, repayment frequency, bank card transfer in and out amount, etc.; customer rewards features include, for example, whether the customer can be contacted, whether the customer commits to repayment, etc.
Step S130, inputting the customer feature library into a preset decision tree algorithm model so that the decision tree algorithm model obtains importance values of the customer features contained in the customer feature library, and selecting a specified number of customer features as candidate customer features according to the order of the importance values from large to small, wherein the importance values are used for representing the importance degree of the customer features relative to asset collection strategy formulation.
The decision tree algorithm is a method for approaching discrete function values, firstly, data are processed, readable rules and decision trees are generated by utilizing a generalization algorithm, and then, new data are analyzed by using decisions. The decision tree is a decision analysis method for evaluating the risk of the project and judging the feasibility of the project by constructing the decision tree to obtain the probability that the expected value of the net present value is greater than or equal to zero on the basis of knowing the occurrence probability of various conditions, and is a graphical method for intuitively applying probability analysis.
The decision tree algorithm model is easy to understand and realize, has the capability of understanding the meaning expressed by the decision tree after interpretation, can process data type and conventional type attributes simultaneously, can make feasible and good-effect results on a large-scale data source in a relatively short time, is insensitive to missing values, can process irrelevant feature data, has the advantages that the decision tree is only required to be built once, the repeated use efficiency is high, and the like, and the client features related to the embodiment can generally obtain better effects through the processing of the decision tree algorithm model, so the embodiment adopts the decision tree algorithm model to process the client features.
According to the method, the importance of the client features relative to the asset collection strategy is converted into specific values through a decision tree algorithm, and then a certain number of client features with larger importance values are screened out to serve as candidate client features. It will be appreciated that the importance value is used to characterize the importance of the customer feature relative to the asset collection policy formulation, so that the candidate customer feature that is screened out is a customer feature that has a strong correlation with the asset collection policy formulation by the bank.
Step S150, inputting the obtained candidate client features into a preset logistic regression algorithm model to obtain a prediction result output by the logistic regression algorithm model, wherein the prediction result comprises a first probability corresponding to a first client type and a second probability corresponding to a second client type.
Logistic regression, also known as logistic regression analysis, is one of the classification and prediction algorithms that predicts the probability of future outcome occurrence through the manifestation of historical data. The logistic regression model is a linear regression normalized by a logistic equation and can be expressed by the following formula:
The meaning of the above formula is: the logarithmic probability of output y=1 is represented by a linear function of input x. It can be seen that as the value of e a+bx approaches positive infinity, the p probability value also approaches 1. The thinking of logistic regression is to fit a decision boundary first and then establish the probability connection between the boundary and classification, so as to obtain the probability under the condition of classification.
The adaptation conditions of logistic regression generally include: the dependent variable is a classified variable of two classifications or the occurrence rate of a certain event, and is a numerical variable; both the residual and dependent variables obey binomial distribution; the argument and the logic probability are linear relationships; the observation objects are mutually independent, and the candidate client characteristics obtained through screening meet the condition, so that the embodiment adopts a logistic regression algorithm model to conduct classification processing on the candidate client characteristics.
The two classification types constructed in the embodiment include a first customer type and a second customer type, wherein the first customer type is used for representing that the customer evaluation is better and corresponds to an asset collection policy for collecting in a bank, and the second customer type is used for representing that the customer evaluation is worse and corresponds to an asset collection policy for collecting the customer entrusted to a third party service company.
According to the embodiment, the obtained candidate client features are input into a preset logistic regression algorithm model, so that a prediction result correspondingly output by the logistic regression algorithm model can be obtained, and a first probability with better client evaluation and a second probability with worse client evaluation can be obtained according to the prediction result.
Step S170, determining the client type with the maximum probability value from the prediction result, and selecting an asset collection strategy corresponding to the client type with the maximum probability value as a target strategy for asset collection for the client.
According to the method and the device, the client type with the largest probability value is determined from the prediction result, and the asset collection strategy corresponding to the client type with the largest probability value is selected as the target strategy for asset collection of the client, so that the strategy for asset collection of the client can be obtained rapidly.
It can be seen from the above that, in the method provided in this embodiment, the intelligent formulation of the customer collecting policy in the asset collecting stage is realized based on the decision tree and logistic regression architecture, the screening of the customer characteristics is realized through the decision tree algorithm model, and then the quality of the customer rating is predicted according to the screened candidate customer characteristics through the logistic regression algorithm model, that is, the asset collecting policy is converted into a two-class problem through the data analysis mode.
Referring to fig. 3, fig. 3 is a flowchart illustrating an artificial intelligence based asset collection method according to another exemplary embodiment of the present application. The asset collection method based on artificial intelligence shown in fig. 3 further includes, based on the steps shown in fig. 2, before step S130:
Step S210, data preprocessing is carried out on the client features contained in the client feature library, so that the client feature library obtained after processing is input into a decision tree algorithm model, and the data preprocessing comprises at least one of data processing and data screening.
The data processing refers to processing data on some customer features in the customer feature library so that the processed customer features better meet the determination of customer evaluation, for example, the data processing includes adding tag information to the refund data of the customer, for example, processing the refund data of the customer into refunds of the customer in about 1 month, refunds in about 3 months, and the like.
The data filtering process refers to data filtering on some client features in the client feature library, and specific modes of data filtering include, but are not limited to, feature filtering or feature replacement, wherein the client features to be filtered include, for example, client features with higher missing values, client features which cannot be used as model input features (such as lack of IDs and names), client features with more abnormal values (such as features with negative amounts), and client features to be replaced use median or average value to replace the client features with abnormal values.
It can be seen that, in this embodiment, the client feature library obtained after the processing is input into the decision tree algorithm model through pre-processing the client features contained in the client feature library in advance, so that the data of the obtained client feature library is more reliable, and the modeling requirement is more met, thereby facilitating the analysis and processing of the subsequent model.
Referring to fig. 4, fig. 4 is a flowchart illustrating an artificial intelligence based asset collection method according to another exemplary embodiment of the present application. The asset collection method based on artificial intelligence shown in fig. 4 further includes, based on the steps shown in fig. 2, after step S130:
And step S310, calculating information values corresponding to the candidate client features, and displaying the obtained calculation result, wherein the information value user characterizes the contribution degree of the candidate client features relative to the asset collection policy.
It should be noted that, the information value is also referred to as IV (Information Value) value, and is mainly used for encoding an input variable and evaluating the prediction capability, where the magnitude of the IV value of the feature variable is indicative of the strength of the prediction capability of the variable.
After obtaining the information values corresponding to the candidate client features, the embodiment further displays the information values corresponding to the candidate client features, for example, sends the information values to a user terminal for display, so that expert users who perform asset collection in banks can check the information values corresponding to the candidate client features, and further select client features from the candidate client features according to the information values corresponding to the candidate client features and self collection experience to input the client features into a logistic regression algorithm model.
Step S330, screening candidate client features from the candidate client features according to the user control information input in real time, so as to input the screened candidate client features into a logistic regression algorithm model, wherein the user control information is used for representing client feature selection operation made by expert users based on self-cleaning experience.
In the embodiment, considering that the clearing opinion of the expert for clearing the assets in the bank can play a certain role in the application scene of clearing the assets, the candidate client features are further screened by combining the information value and the opinion of the expert user.
According to the method, candidate client features are filtered from the candidate client features according to the user control information input in real time, so that the filtered candidate client features are input into a logistic regression algorithm model, wherein the user control information input in real time is the user control information which is triggered and input in real time in a user terminal for displaying information values corresponding to the candidate client features, and the user control information is used for representing client feature selection operations made by expert users based on self-cleaning experience. For example, after the information value corresponding to each candidate client feature is presented to the expert user, the expert user determines, in combination with the information value, whether the corresponding candidate client feature has a great influence on the asset clearance, for example, whether the information value corresponding to the candidate client feature is "available" is not great, but the expert considers that the candidate client feature has a great influence on the asset clearance in practical application, so that the expert user manually selects to input the candidate client feature into the logistic regression algorithm model, and in addition, the expert user also selects to input some candidate client features with great information values into the logistic regression algorithm model. From this, it can be derived that the present embodiment combines the actual experience of the expert to promote the accuracy of the asset collection policy selection.
Referring to fig. 5, fig. 5 is a flowchart of step S310 in the embodiment shown in fig. 4 in an exemplary embodiment. As shown in fig. 5, the information value corresponding to each candidate client feature can be calculated through steps S311 to S315, which will be described in detail below:
Step S311, respectively performs grouping processing on each candidate client feature to obtain a plurality of feature groups corresponding to each candidate client feature.
The process of grouping one of the candidate client features is essentially a process of discretizing the candidate client feature, and thus the candidate client feature is typically a continuous type of feature. For example, the candidate client features may be classified into a plurality of feature groups by performing feature binning on the candidate client features, and specific feature binning methods include, but are not limited to, supervised chi-square binning (ChiMerge), equidistant binning, and equal-frequency binning, which may be selected according to the data condition of the candidate client features in an actual application scenario, which is not limited in this embodiment.
Step S313, calculating information values corresponding to each feature group based on the feature information distribution in each feature group.
In this embodiment, the information value corresponding to each feature group may be calculated by the following formula:
Wherein IV i represents the information value corresponding to the ith feature group of candidate client features, WOE i represents evidence weight, meaning the difference between "the proportion of feature yi of responding client in the current group to feature ys of all responding clients" and "the proportion of feature ni of no responding client in the current group to feature ns of all no responding client" is understood to mean that the feature of responding client contributes to the result of evaluating the client as a better type, and similarly, the feature of no responding client is understood to mean that the feature of not contributing to the result of evaluating the client as a better type; n represents the number of feature packets; (P yi-Pni) represents the difference between the "proportion of the features of the responding client in the current packet to the features of all responding clients" and the "proportion of the features of the non-responding client in the current packet to the features of all non-responding clients", thereby ensuring that the information value is not negative.
From this, it can be seen that the information value represents the proportion of the number of individuals in the current group of variables to the whole, so that the proportion of the samples in the group to the whole is well considered, and the lower the proportion is, the lower the contribution of the group to the whole prediction capability of the variables is, namely the lower the contribution degree to the asset collection policy is.
Step S315, calculating the information values corresponding to the candidate client features according to the information values corresponding to all the feature groups.
The process of calculating the information values corresponding to the candidate client features according to the information values corresponding to all feature groups in this embodiment can be expressed by the following formula:
it can be seen that, in this embodiment, the sum of the information values corresponding to all the feature groups is calculated, and the obtained sum is used as the information value corresponding to the corresponding candidate client feature.
According to the method, the candidate client features are grouped, the information value corresponding to each feature group is calculated, and finally the root takes the sum of the information values corresponding to all the feature groups as the information value corresponding to the corresponding candidate client features, so that the contribution degree of the candidate client features to the asset collection strategy can be accurately evaluated by the obtained information values, and the candidate client features are accurately screened.
Referring to FIG. 6, FIG. 6 is a flow chart illustrating an artificial intelligence based asset collection method according to another exemplary embodiment of the application. As shown in fig. 6, before step S150 in the embodiment shown in fig. 2, the method further includes steps S410 to S450, which are described in detail below:
In step S410, the asset collection data handed over in the first time period is selected as the training sample set from the historical asset collection data, and the asset collection data handed over in the second time period is selected as the verification sample set, wherein the first time period is earlier than the second time period.
This embodiment describes the process of training to obtain a logistic regression algorithm model. The historical asset collection data refers to data generated by a sample customer asset after being handed over to the management department of the special management asset within the bank, including, for example, customer generated repayment behaviors, corresponding asset information of the customer after the node, liability information, credit assessment data, and the like.
In this embodiment, the asset collection data intersected in the previous time period is used as a training sample set, the asset collection data intersected in the later time period is used as a verification sample set, the training sample set and the verification sample set both comprise a plurality of positive samples and a plurality of negative samples, the positive samples contribute to the better result of evaluating the client, and the negative samples contribute to the worse result of evaluating the client. For example, the positive sample is the clearance data of more than 5% of customers who pay more than 5% in 5 months handed over to the management department, and the negative sample is the clearance data of less than 5% of customers who pay less than 5% in 5 months handed over to the management department.
Step S430, training the logistic regression algorithm model to be trained according to the training sample set, and verifying the trained logistic regression algorithm model through the verification sample set to obtain a verification result.
The process of training the logistic regression algorithm model to be trained according to the training sample set and then verifying the trained logistic regression algorithm model through the verification sample set comprises the following steps: obtaining a prediction result output by the trained logistic regression algorithm model aiming at each verification sample in the verification sample set; and calculating a model effect evaluation value according to the obtained prediction result, and if the model effect evaluation value is greater than or equal to a preset threshold value, verifying that the trained logistic regression algorithm model meets the requirement, and generating a verification result for indicating that the trained logistic regression algorithm model meets the requirement. Otherwise, if the model effect evaluation value is smaller than the preset threshold value, verifying that the trained logistic regression algorithm model does not meet the requirement, and generating a verification result for indicating that the trained logistic regression algorithm model does not meet the requirement.
The model effect evaluation value includes an AUC (Area Underner Curve, defined as the Area enclosed by the coordinate axis Under the ROC Curve) value or a KS (Kolmogorov-Smirnov, which is the difference between the accumulated parts of the good and bad samples, and represents the ability of the model to distinguish between the positive and negative samples), and the larger the AUC value or KS value, the better the model effect is.
Table 1 below is an example of determining the model effect of a logistic regression algorithm model by AUC or KS values:
TABLE 1
Step S450, if the verification result indicates that the trained logistic regression algorithm model meets the requirements, the candidate client features are input into the trained logistic regression algorithm model.
And if the trained logistic regression algorithm model obtained through verification in the steps meets the requirements, inputting the candidate client features into the trained logistic regression algorithm model. In an exemplary application scenario, if the predicted variable is "in-bank revenue score", the modeling variable includes the client features "active client flag", "delinquent rebate total", "current overdue days", "default amount total", "average refund amount of last three credit months" and "last refund date from current days", and the logistic regression equation based on the training sample and the verification sample may be: ln (p/1-p) =0.22+2.26 active customer flag+0.26 delinquent compound total+0.09 current overdue day+0.26 default total+0.50 average refund of last three credit months+0.34 last refund date from current day. The preset parameters corresponding to the variables are model parameters obtained through training and verification.
And if the verification result indicates that the trained logistic regression algorithm model does not meet the requirements, repeating the steps of training the logistic regression algorithm model to be trained according to the training sample set, verifying the trained logistic regression algorithm model through the verification sample set to obtain the verification result until the obtained verification result indicates that the trained logistic regression algorithm model meets the requirements.
Therefore, the logistic regression algorithm model is trained through the method of the embodiment, so that the trained logistic regression algorithm model can be suitable for the application scene of asset collection, the classification probability value predicted by the logistic regression algorithm model subsequently has high accuracy, and the method has great contribution to the actual application scene.
FIG. 7 is a block diagram of an asset collection device based on artificial intelligence, as shown in an exemplary embodiment of the application. As shown in fig. 7, the apparatus includes:
A feature acquisition module 510 configured to acquire a customer feature library associated with an asset to be cleared; the feature selection module 530 is configured to input the client feature library into a preset decision tree algorithm model, so that the decision tree algorithm model obtains importance values of client features contained in the client feature library, and selects a specified number of client features as candidate client features according to the order of the importance values from large to small, wherein the importance values are used for representing importance degrees of the client features relative to asset collection policy formulation; the probability prediction module 550 is configured to input the obtained candidate client features into a preset logistic regression algorithm model to obtain a prediction result output by the logistic regression algorithm model, wherein the prediction result comprises a first probability corresponding to a first client type and a second probability corresponding to a second client type; the policy selection module 570 is configured to determine a client type with the largest probability value from the prediction result, and select an asset collection policy corresponding to the client type with the largest probability value as a target policy for asset collection for the client.
The device is based on a decision tree and logistic regression architecture to realize the intelligent formulation of the customer clearing strategy in the asset clearing stage, and compared with the prior art that the customer clearing strategy is selected by subjective evaluation by professionals responsible for clearing in banks, the technical scheme provided by the application can avoid artificial subjective influence and rely on the processing of customer characteristics related to the asset clearing, so that the obtained asset clearing strategy is more reliable and has higher efficiency.
In another exemplary embodiment, the apparatus further comprises:
The information value acquisition module is configured to calculate information values corresponding to the candidate client features, and display the obtained calculation results, wherein the information value user characterizes the contribution degree of the candidate client features relative to the asset collection strategy; and the feature screening module is configured to screen candidate client features from the candidate client features according to user control information input in real time so as to input the screened candidate client features into the logistic regression algorithm model, wherein the user control information is used for representing client feature selection operation of expert users based on self-cleaning experience.
In another exemplary embodiment, the information value acquisition module includes:
The feature grouping unit is configured to perform grouping processing on each candidate client feature respectively so as to obtain a plurality of feature groups corresponding to each candidate client feature respectively; a feature group calculation unit configured to calculate information values corresponding to each feature group based on feature information distribution within each feature group, respectively; and the comprehensive calculation unit is configured to calculate the information values corresponding to the candidate client features according to the information values corresponding to all the feature groups.
In another exemplary embodiment, the apparatus further comprises:
The preprocessing module is configured to perform data preprocessing on the client features contained in the client feature library so as to input the client feature library obtained after the processing into the decision tree algorithm model, wherein the data preprocessing comprises at least one of data processing and data screening.
In another exemplary embodiment, the apparatus further comprises:
the sample set acquisition module is configured to select asset collection data handed over in a first time period from the historical asset collection data as a training sample set, and select asset collection data handed over in a second time period as a verification sample set, wherein the first time period is earlier than the second time period; the model training and verifying module is configured to train the logistic regression algorithm model to be trained according to the training sample set, and verify the trained logistic regression algorithm model through the verifying sample set to obtain a verification result; and the feature input module is configured to input the candidate client features into the trained logistic regression algorithm model if the verification result indicates that the trained logistic regression algorithm model meets the requirements.
In another exemplary embodiment, the apparatus further comprises:
And the circulation module is configured to repeatedly execute the steps of training the logistic regression algorithm model to be trained according to the training sample set and verifying the trained logistic regression algorithm model through the verification sample set to obtain a verification result if the verification result indicates that the trained logistic regression algorithm model does not meet the requirements, until the obtained verification result indicates that the trained logistic regression algorithm model meets the requirements.
In another exemplary embodiment, the model training and verification module includes:
The training result acquisition unit is configured to acquire the prediction result output by the trained logistic regression algorithm model aiming at each verification sample in the verification sample set; and the verification result acquisition unit is configured to calculate a model effect evaluation value according to the acquired prediction result, and if the model effect evaluation value is greater than a preset threshold value, a verification result indicating that the trained logistic regression algorithm model meets the requirement is generated.
It should be noted that, the apparatus provided in the foregoing embodiments and the method provided in the foregoing embodiments belong to the same concept, and the specific manner in which each module and unit perform the operation has been described in detail in the method embodiments, which is not repeated herein.
The embodiment of the application also provides electronic equipment, which comprises: one or more processors; and a storage device for storing one or more programs which, when executed by the one or more processors, cause the electronic device to implement the artificial intelligence based asset collection method described in the foregoing embodiments.
Fig. 8 shows a schematic diagram of a computer system suitable for use in implementing an embodiment of the application. It should be noted that, the computer system 1600 of the electronic device shown in fig. 8 is only an example, and should not impose any limitation on the functions and the application scope of the embodiments of the present application.
As shown in fig. 8, the computer system 1600 includes a central processing unit (Central Processing Unit, CPU) 1601 that can perform various appropriate actions and processes, such as performing the methods described in the above embodiments, according to a program stored in a Read-Only Memory (ROM) 1602 or a program loaded from a storage portion 1608 into a random access Memory (Random Access Memory, RAM) 1603. In the RAM 1603, various programs and data required for system operation are also stored. The CPU 1601, ROM 1602, and RAM 1603 are connected to each other by a bus 1604. An Input/Output (I/O) interface 1605 is also connected to bus 1604.
The following components are connected to the I/O interface 1605: an input portion 1606 including a keyboard, a mouse, and the like; an output portion 1607 including a Cathode Ray Tube (CRT), a Liquid crystal display (Liquid CRYSTAL DISPLAY, LCD), and a speaker, etc.; a storage portion 1608 including a hard disk or the like; and a communication section 1609 including a network interface card such as a LAN (Local Area Network ) card, a modem, or the like. The communication section 1609 performs communication processing via a network such as the internet. The drive 1610 is also connected to the I/O interface 1605 as needed. A removable medium 1611 such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, or the like is installed as needed on the drive 1610 so that a computer program read out therefrom is installed into the storage section 1608 as needed.
In particular, according to embodiments of the present application, the processes described above with reference to flowcharts may be implemented as computer software programs. For example, embodiments of the present application include a computer program product comprising a computer program embodied on a computer readable medium, the computer program comprising a computer program for performing the method shown in the flowchart. In such embodiments, the computer program may be downloaded and installed from a network via the communication portion 1609, and/or installed from the removable media 1611. When executed by a Central Processing Unit (CPU) 1601, performs various functions defined in the system of the present application.
It should be noted that, the computer readable medium shown in the embodiments of the present application may be a computer readable signal medium or a computer readable storage medium, or any combination of the two. The computer readable storage medium may be, for example, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of the computer-readable storage medium may include, but are not limited to: an electrical connection having one or more wires, a portable computer diskette, a hard disk, a Random Access Memory (RAM), a read-Only Memory (ROM), an erasable programmable read-Only Memory (Erasable Programmable Read Only Memory, EPROM), a flash Memory, an optical fiber, a portable compact disc read-Only Memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing. In the context of this document, a computer readable storage medium may be any tangible medium that can contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device. In the present application, however, a computer-readable signal medium may include a data signal propagated in baseband or as part of a carrier wave, with a computer-readable computer program embodied therein. Such a propagated data signal may take any of a variety of forms, including, but not limited to, electro-magnetic, optical, or any suitable combination of the foregoing. A computer readable signal medium may also be any computer readable medium that is not a computer readable storage medium and that can communicate, propagate, or transport a program for use by or in connection with an instruction execution system, apparatus, or device. A computer program embodied on a computer readable medium may be transmitted using any appropriate medium, including but not limited to: wireless, wired, etc., or any suitable combination of the foregoing.
The flowcharts and block diagrams in the figures illustrate the architecture, functionality, and operation of possible implementations of systems, methods and computer program products according to various embodiments of the present application. Where each block in the flowchart or block diagrams may represent a module, segment, or portion of code, which comprises one or more executable instructions for implementing the specified logical function(s). It should also be noted that, in some alternative implementations, the functions noted in the block may occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently, or the blocks may sometimes be executed in the reverse order, depending upon the functionality involved. It will also be noted that each block of the block diagrams or flowchart illustration, and combinations of blocks in the block diagrams or flowchart illustration, can be implemented by special purpose hardware-based systems which perform the specified functions or acts, or combinations of special purpose hardware and computer instructions.
The units involved in the embodiments of the present application may be implemented by software, or may be implemented by hardware, and the described units may also be provided in a processor. Wherein the names of the units do not constitute a limitation of the units themselves in some cases.
Another aspect of the application also provides a computer readable storage medium having stored thereon a computer program which when executed by a processor implements an artificial intelligence based asset collection method as described above. The computer-readable storage medium may be included in the electronic device described in the above embodiment or may exist alone without being incorporated in the electronic device.
The foregoing is merely illustrative of the preferred embodiments of the present application and is not intended to limit the embodiments of the present application, and those skilled in the art can easily make corresponding variations or modifications according to the main concept and spirit of the present application, so that the protection scope of the present application shall be defined by the claims.
Claims (8)
1. An asset collection method based on artificial intelligence, which is characterized by comprising the following steps:
Acquiring a customer feature library associated with the asset to be cleared;
Inputting the client feature library into a preset decision tree algorithm model, so that the decision tree algorithm model obtains importance values of client features contained in the client feature library, and selecting a specified number of client features as candidate client features according to the sequence of the importance values from large to small, wherein the importance values are used for representing importance degrees of the client features relative to asset collection strategy formulation;
Inputting the obtained candidate client characteristics into a preset logistic regression algorithm model to obtain a prediction result output by the logistic regression algorithm model, wherein the prediction result comprises a first probability corresponding to a first client type and a second probability corresponding to a second client type, the client evaluation of the first client type is superior to the client evaluation of the second client type, the first client type corresponds to an asset collection policy for promoting collection in a bank, and the second client type corresponds to an asset collection policy for entrusting a client to a third party service company for collection;
Determining a client type with the maximum probability value from the prediction result, and selecting an asset collection strategy corresponding to the client type with the maximum probability value as a target strategy for asset collection for the client;
After inputting the client feature library into a preset decision tree algorithm model so that the decision tree algorithm model obtains importance values of client features contained in the client feature library, selecting a specified number of client features as candidate client features according to the order of the importance values from large to small, wherein the method further comprises the steps of: calculating information values corresponding to the candidate client features, and displaying the obtained calculation results, wherein the information value user represents the contribution degree of the candidate client features relative to the asset collection strategy;
Screening candidate client features from the candidate client features according to user control information input in real time to input the screened candidate client features into the logistic regression algorithm model, wherein the user control information is used for representing client feature selection operations made by expert users based on self-cleaning experience;
calculating the information value corresponding to each candidate client feature, including:
grouping processing is carried out on each candidate client feature respectively so as to obtain a plurality of feature groups corresponding to each candidate client feature respectively;
calculating information values corresponding to each feature group based on feature information distribution in each feature group;
And calculating the information values corresponding to the candidate client features according to the information values corresponding to all the feature groups.
2. The method of claim 1, wherein before inputting the customer feature library into a preset decision tree algorithm model to cause the decision tree algorithm model to obtain importance values for customer features contained in the customer feature library, selecting a specified number of customer features as candidate customer features in a ranking of importance values from large to small, the method further comprises:
And carrying out data preprocessing on the client features contained in the client feature library so as to input the client feature library obtained after the processing into the decision tree algorithm model, wherein the data preprocessing comprises at least one of data processing and data screening.
3. The method of claim 1, wherein prior to inputting the resulting candidate customer features into a pre-set logistic regression algorithm model, the method further comprises:
Selecting asset collection data handed over in a first time period from historical asset collection data as a training sample set, and selecting asset collection data handed over in a second time period as a verification sample set, wherein the first time period is earlier than the second time period;
training the logistic regression algorithm model to be trained according to the training sample set, and verifying the trained logistic regression algorithm model through the verification sample set to obtain a verification result;
And if the verification result indicates that the trained logistic regression algorithm model meets the requirement, inputting the candidate client features into the trained logistic regression algorithm model.
4. A method according to claim 3, characterized in that the method further comprises:
And if the verification result indicates that the trained logistic regression algorithm model does not meet the requirements, repeating the steps of training the logistic regression algorithm model to be trained according to the training sample set, verifying the trained logistic regression algorithm model through the verification sample set to obtain the verification result until the obtained verification result indicates that the trained logistic regression algorithm model meets the requirements.
5. The method according to claim 3 or 4, wherein verifying the trained logistic regression algorithm model through the verification sample set to obtain a verification result comprises:
obtaining a prediction result output by the trained logistic regression algorithm model aiming at each verification sample in the verification sample set;
And calculating a model effect evaluation value according to the obtained prediction result, and if the model effect evaluation value is larger than a preset threshold value, generating a verification result for indicating that the trained logistic regression algorithm model meets the requirement.
6. An artificial intelligence based asset collection device, comprising:
the feature acquisition module is configured to acquire a customer feature library associated with the asset to be cleared;
The feature selection module is configured to input the customer feature library into a preset decision tree algorithm model, so that the decision tree algorithm model obtains importance values of the customer features contained in the customer feature library, and a specified number of customer features are selected as candidate customer features according to the sequence of the importance values from large to small, wherein the importance values are used for representing importance degrees of the customer features formulated relative to an asset collection strategy;
The probability prediction module is configured to input the obtained candidate client characteristics into a preset logistic regression algorithm model to obtain a prediction result output by the logistic regression algorithm model, wherein the prediction result comprises a first probability corresponding to a first client type and a second probability corresponding to a second client type, the client evaluation of the first client type is superior to the client evaluation of the second client type, the first client type corresponds to an asset collection policy for collecting in a bank, and the second client type corresponds to an asset collection policy for collecting clients delegated to a third party service company;
The strategy selection module is configured to determine the client type with the maximum probability value from the prediction result, and select an asset collection strategy corresponding to the client type with the maximum probability value as a target strategy for asset collection for the client;
the apparatus further comprises: the information value acquisition module is configured to calculate information values corresponding to the candidate client features, and display the obtained calculation results, wherein the information value user characterizes the contribution degree of the candidate client features relative to the asset collection strategy;
The feature screening module is configured to screen candidate client features from the candidate client features according to user control information input in real time so as to input the screened candidate client features into the logistic regression algorithm model, wherein the user control information is used for representing client feature selection operation of expert users based on self-cleaning experience;
the information value acquisition module includes: the feature grouping unit is configured to perform grouping processing on each candidate client feature respectively so as to obtain a plurality of feature groups corresponding to each candidate client feature respectively; a feature group calculation unit configured to calculate information values corresponding to each feature group based on feature information distribution within each feature group, respectively; and the comprehensive calculation unit is configured to calculate the information values corresponding to the candidate client features according to the information values corresponding to all the feature groups.
7. An electronic device, comprising:
one or more processors;
storage means for storing one or more programs that, when executed by the one or more processors, cause the electronic device to implement the artificial intelligence based asset collection method of any of claims 1 to 5.
8. A computer readable storage medium having stored thereon computer readable instructions which, when executed by a processor of a computer, cause the computer to perform the artificial intelligence based asset collection method of any of claims 1 to 5.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111155958.4A CN113822490B (en) | 2021-09-29 | 2021-09-29 | Asset collection method and device based on artificial intelligence and electronic equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111155958.4A CN113822490B (en) | 2021-09-29 | 2021-09-29 | Asset collection method and device based on artificial intelligence and electronic equipment |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113822490A CN113822490A (en) | 2021-12-21 |
| CN113822490B true CN113822490B (en) | 2024-05-14 |
Family
ID=78916063
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111155958.4A Active CN113822490B (en) | 2021-09-29 | 2021-09-29 | Asset collection method and device based on artificial intelligence and electronic equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113822490B (en) |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108256691A (en) * | 2018-02-08 | 2018-07-06 | 成都智宝大数据科技有限公司 | Refund Probabilistic Prediction Model construction method and device |
| CN110659318A (en) * | 2019-08-15 | 2020-01-07 | 中国平安财产保险股份有限公司 | Big data based strategy pushing method and system and computer equipment |
| CN112200583A (en) * | 2020-10-28 | 2021-01-08 | 交通银行股份有限公司 | Knowledge graph-based fraud client identification method |
| CN113112347A (en) * | 2021-05-07 | 2021-07-13 | 中国建设银行股份有限公司 | Determination method of hasty collection decision, related device and computer storage medium |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US9792567B2 (en) * | 2016-03-11 | 2017-10-17 | Route4Me, Inc. | Methods and systems for managing large asset fleets through a virtual reality interface |
-
2021
- 2021-09-29 CN CN202111155958.4A patent/CN113822490B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108256691A (en) * | 2018-02-08 | 2018-07-06 | 成都智宝大数据科技有限公司 | Refund Probabilistic Prediction Model construction method and device |
| CN110659318A (en) * | 2019-08-15 | 2020-01-07 | 中国平安财产保险股份有限公司 | Big data based strategy pushing method and system and computer equipment |
| CN112200583A (en) * | 2020-10-28 | 2021-01-08 | 交通银行股份有限公司 | Knowledge graph-based fraud client identification method |
| CN113112347A (en) * | 2021-05-07 | 2021-07-13 | 中国建设银行股份有限公司 | Determination method of hasty collection decision, related device and computer storage medium |
Non-Patent Citations (1)
| Title |
|---|
| LightGBM算法在早期催收管理工作中的应用;王方春;;电脑知识与技术;20200305(07);211-212+227 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113822490A (en) | 2021-12-21 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Kozodoi et al. | Fairness in credit scoring: Assessment, implementation and profit implications | |
| CN113240509B (en) | Loan risk assessment method based on multi-source data federal learning | |
| KR20210116439A (en) | Systems and Methods for Anti-Money Laundering Analysis | |
| JP2020522832A (en) | System and method for issuing a loan to a consumer determined to be creditworthy | |
| Höglund | Tax payment default prediction using genetic algorithm-based variable selection | |
| US11538044B2 (en) | System and method for generation of case-based data for training machine learning classifiers | |
| CN112633962B (en) | Service recommendation method and device, computer equipment and storage medium | |
| CN110909984A (en) | Business data processing model training method, business data processing method and device | |
| CN113609193A (en) | Method and device for training prediction model for predicting customer transaction behavior | |
| CN111192133A (en) | Method and device for generating risk model after user loan and electronic equipment | |
| CN114078050A (en) | Loan overdue prediction method and device, electronic equipment and computer readable medium | |
| Eddy et al. | Credit scoring models: Techniques and issues | |
| CN112232950A (en) | Loan risk assessment method and device, equipment and computer-readable storage medium | |
| CN115545886A (en) | Overdue risk identification method, overdue risk identification device, overdue risk identification equipment and storage medium | |
| CN112200684B (en) | Method, system and storage medium for detecting medical insurance fraud | |
| CN114493686A (en) | Operation content generation and pushing method and device | |
| KR102499181B1 (en) | Loan regular auditing system using artificia intellicence | |
| CN113822490B (en) | Asset collection method and device based on artificial intelligence and electronic equipment | |
| CN111160647A (en) | Money laundering behavior prediction method and device | |
| Muzammil et al. | Determinants for the Adoption of Regulatory Technology (RegTech) Services by the Companies in United Arab Emirates: An MCDM Approach. | |
| CN114170000A (en) | Credit card user risk category identification method, device, computer equipment and medium | |
| Bagde et al. | Analysis of fraud detection mechanism in health insurance using statistical data mining techniques | |
| Cun | Business simulation analysis based on leadership development with crisis management using gaming techniques and machine learning model | |
| Yazdani | Developing a model for validation and prediction of bank customer credit using information technology (case study of Dey Bank) | |
| CN113032643B (en) | Target behavior recognition system |
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 |