CN115409635A - Information prediction method, device, equipment and medium - Google Patents

Information prediction method, device, equipment and medium Download PDF

Info

Publication number
CN115409635A
CN115409635A CN202211068161.5A CN202211068161A CN115409635A CN 115409635 A CN115409635 A CN 115409635A CN 202211068161 A CN202211068161 A CN 202211068161A CN 115409635 A CN115409635 A CN 115409635A
Authority
CN
China
Prior art keywords
sample data
information
classification
iterative training
training
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN202211068161.5A
Other languages
Chinese (zh)
Inventor
李明骏
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Industrial and Commercial Bank of China Ltd ICBC
Original Assignee
Industrial and Commercial Bank of China Ltd ICBC
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Industrial and Commercial Bank of China Ltd ICBC filed Critical Industrial and Commercial Bank of China Ltd ICBC
Priority to CN202211068161.5A priority Critical patent/CN115409635A/en
Publication of CN115409635A publication Critical patent/CN115409635A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q40/00—Finance; Insurance; Tax strategies; Processing of corporate or income taxes
    • G06Q40/08—Insurance

Landscapes

  • Business, Economics & Management (AREA)
  • Accounting & Taxation (AREA)
  • Finance (AREA)
  • Engineering & Computer Science (AREA)
  • Development Economics (AREA)
  • Economics (AREA)
  • Marketing (AREA)
  • Strategic Management (AREA)
  • Technology Law (AREA)
  • Physics & Mathematics (AREA)
  • General Business, Economics & Management (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

The disclosure provides an information prediction method, an information prediction device, information prediction equipment and an information prediction medium, which can be applied to the technical field of big data and the technical field of finance. The information prediction method comprises the following steps: the method comprises the steps that operation condition information of a target enterprise is obtained from a database in response to a credit risk prediction request for the target enterprise from a client; inputting the operation condition information into a classification model and outputting a classification result; the classification model is obtained by initializing sample data by using chaotic mapping to obtain a sample data set and then performing iterative training on a preset model, wherein the sample data comprises operation condition information and credit category information of an enterprise, and model parameters of the t +1 th iterative training are determined according to the result of the t th iterative training, wherein t is a positive integer; and generating credit risk prediction result information of the target enterprise according to the classification result.

Description

Information prediction method, device, equipment and medium
Technical Field
The present disclosure relates to the field of big data technology and financial technology, and in particular, to an information prediction method, apparatus, device, medium, and program product.
Background
The credit risk assessment work is an important link in the loan approval process of the financial institution, and has important influence on the approval result of the loan approval business. With the advent of the big data era, when a financial institution evaluates credit risk of an enterprise, the types and the amount of data to be faced are more and more, and the relevance between the data is stronger, so that the result accuracy of the traditional credit risk evaluation method on the credit risk evaluation is lower.
Disclosure of Invention
In view of the foregoing, the present disclosure provides information prediction methods, apparatuses, devices, media, and program products that improve the accuracy of credit risk assessment.
According to an aspect of the present disclosure, there is provided an information prediction method including:
the method comprises the steps that operation condition information of a target enterprise is obtained from a database in response to a credit risk prediction request for the target enterprise from a client;
inputting the operation condition information into a classification model, and outputting a classification result; the classification model is obtained by initializing sample data by utilizing chaotic mapping to obtain a sample data set and then performing iterative training on a preset model, wherein the sample data comprises the operation condition information and the credit category information of an enterprise, and the model parameter of the t +1 iterative training is determined according to the result of the t iterative training, wherein t is a positive integer; and
and generating credit risk prediction result information of the target enterprise according to the classification result.
According to the embodiment of the disclosure, the training method of the classification model comprises the following steps:
acquiring operation condition information and credit category information of n enterprises as sample data, wherein the operation condition information of each enterprise comprises m operation index information, and m and n are positive integers;
determining iteration conditions and initialization parameters according to application requirements of a credit risk prediction scene;
initializing sample data by utilizing chaotic mapping according to the initialization parameters to obtain a sample data set;
and inputting the sample data set into a preset model for iterative training according to the iterative conditions to obtain a trained classification model.
According to the embodiment of the disclosure, inputting a sample data set into a preset model for iterative training to obtain a trained classification model;
inputting the sample data set into a preset model, and outputting a classification result, wherein the classification result represents the fitness of each sample data in the sample data set;
determining model parameters of the t +1 th iterative training by adopting a self-adaptive weight method according to the fitness of the sample data in the t-th iterative training;
and performing iterative training on the preset model by adopting a k-fold cross validation strategy until the classification result meets an iterative condition to obtain a trained classification model, wherein k is a positive integer which is more than 5 and less than 20.
According to the embodiment of the disclosure, iterative training is performed on a preset model by adopting a k-fold cross validation strategy until a classification result meets an iterative condition, so as to obtain a trained classification model, which includes:
for the t-th iterative training, dividing a sample data set into k groups, taking the kth group of sample data as a test data set, and taking other k-1 groups of sample data as a training data set;
inputting k groups of sample data into a classification model obtained by training k-1 groups of sample data, and outputting a classification result;
under the condition that all sample data in the sample data set are tested, determining the classification accuracy of the t-th iterative training according to all classification results and credit category information of the t-th iterative training;
and obtaining a trained classification model under the condition that the classification accuracy meets the iteration condition.
According to an embodiment of the present disclosure, the training method of the classification model further includes:
determining the incidence relation of the m pieces of operation index information according to the m pieces of operation index information aiming at the operation condition information of each enterprise;
determining the accumulated contribution rate of each operation index by a principal component analysis method according to the incidence relation;
and performing dimensionality reduction on the m pieces of operation index information according to the accumulated contribution rate to obtain p pieces of operation index information, wherein p is a positive integer smaller than m.
According to an embodiment of the present disclosure, the training method of the classification model further includes:
determining operation trend information of the target enterprise according to the operation condition information of the target enterprise;
and generating credit risk prediction result information of the target enterprise according to the classification result and the operation trend information of the target enterprise.
Another aspect of the present disclosure provides an information prediction apparatus including: the device comprises an acquisition module, a prediction module and a first generation module. The acquisition module is used for responding to a credit risk prediction request aiming at the target enterprise from the client and acquiring the operation condition information of the target enterprise from the database. The prediction module is used for inputting the operation condition information into the classification model and outputting a classification result; the classification model is obtained by initializing sample data by using chaotic mapping to obtain a sample data set and then performing iterative training on a preset model, wherein the sample data comprises operation condition information and credit category information of an enterprise, and model parameters of the t +1 th iterative training are determined according to the result of the t th iterative training, wherein t is a positive integer. And the first generation module is used for generating credit risk prediction result information of the target enterprise according to the classification result.
According to an embodiment of the present disclosure, the prediction module includes an acquisition sub-module, a first determination sub-module, an initialization sub-module, and a training sub-module. The obtaining submodule is used for obtaining the operation condition information and the credit category information of n enterprises as sample data, wherein the operation condition information of each enterprise comprises m pieces of operation index information, and m and n are positive integers. And the first determining submodule is used for determining iteration conditions and initialization parameters according to the application requirements of the credit risk prediction scene. And the initialization module is used for initializing the sample data by utilizing chaotic mapping according to the initialization parameters to obtain the sample data set. And the training submodule is used for inputting the sample data set into a preset model according to the iteration condition to carry out iteration training to obtain a trained classification model.
According to an embodiment of the present disclosure, a training submodule includes an output unit, a determination unit, and a training unit. The output unit is used for inputting the sample data set into a preset model and outputting a classification result, wherein the classification result represents the fitness of each sample data in the sample data set. And the determining unit is used for determining the model parameters of the t +1 th iterative training by adopting a self-adaptive weight method according to the fitness of the sample data in the t-th iterative training. And the training unit is used for performing iterative training on the preset model by adopting a k-fold cross validation strategy until the classification result meets an iterative condition to obtain a trained classification model, wherein k is a positive integer which is more than 5 and less than 20.
According to an embodiment of the present disclosure, a training unit includes a grouping subunit, an output subunit, a first determination subunit, and a second determination subunit. The grouping subunit is configured to, for the t-th iterative training, divide the sample data set into k groups, use the kth group of sample data as a test data set, and use the other k — 1 groups of sample data as a training data set. And the output subunit is used for inputting the k groups of sample data into a classification model obtained by training the k-1 groups of sample data and outputting a classification result. And the first determining subunit is used for determining the classification accuracy of the t-th iterative training according to all classification results and credit category information of the t-th iterative training under the condition that all sample data in the sample data set are tested. And the second determining subunit is used for obtaining the trained classification model under the condition that the classification accuracy rate meets the iteration condition.
According to an embodiment of the present disclosure, the prediction module further includes a second determining sub-module, a third determining sub-module, and a processing sub-module. And the second determining submodule is used for determining the incidence relation of the m pieces of operation index information according to the m pieces of operation index information aiming at the operation condition information of each enterprise. And the third determining submodule is used for determining the accumulated contribution rate of each operation index through a principal component analysis method according to the association relation. And the processing submodule is used for performing dimensionality reduction processing on the m pieces of operation index information according to the accumulated contribution rate to obtain p pieces of operation index information, wherein p is a positive integer smaller than m.
According to an embodiment of the present disclosure, the information prediction apparatus further includes a determining module and a second generating module. The determining module is used for determining the operation trend information of the target enterprise according to the operation condition information of the target enterprise. And the second generation module is used for generating credit risk prediction result information of the target enterprise according to the classification result and the operation trend information of the target enterprise.
Another aspect of the present disclosure provides an electronic device including: one or more processors; a memory for storing one or more programs, wherein the one or more programs, when executed by the one or more processors, cause the one or more processors to perform the above-described information prediction method.
Another aspect of the present disclosure also provides a computer-readable storage medium having stored thereon executable instructions that, when executed by a processor, cause the processor to perform the above-described information prediction method.
Another aspect of the present disclosure also provides a computer program product comprising a computer program which, when executed by a processor, implements the above-described information prediction method.
According to the embodiment of the disclosure, the operation condition information of the target enterprise is acquired from the database, the operation condition information is input into the classification model, the classification result is output, and the credit risk prediction result information of the target enterprise is generated according to the classification result. The classification model is obtained by chaotic mapping and initializing sample data to obtain a sample data set and then performing iterative training on the preset model, so that the optimization efficiency and the traversal uniformity of the input operation condition information are improved, and the accuracy of the credit risk prediction result is improved to a certain extent.
Drawings
The foregoing and other objects, features and advantages of the disclosure will be apparent from the following description of embodiments of the disclosure, which proceeds with reference to the accompanying drawings, in which:
fig. 1 schematically illustrates an application scenario diagram of an information prediction method, apparatus, device, medium, and program product according to embodiments of the present disclosure;
FIG. 2 schematically illustrates a flow diagram of an information prediction method according to an embodiment of the disclosure;
FIG. 3 schematically shows a flow chart of a method of training a classification model according to an embodiment of the present disclosure;
fig. 4 schematically shows a block diagram of the structure of an information prediction apparatus according to an embodiment of the present disclosure; and
fig. 5 schematically shows a block diagram of an electronic device adapted to implement the information prediction method according to an embodiment of the present disclosure.
Detailed Description
Hereinafter, embodiments of the present disclosure will be described with reference to the accompanying drawings. It should be understood that the description is illustrative only and is not intended to limit the scope of the present disclosure. In the following detailed description, for purposes of explanation, numerous specific details are set forth in order to provide a thorough understanding of the embodiments of the disclosure. It may be evident, however, that one or more embodiments may be practiced without these specific details. Moreover, in the following description, descriptions of well-known structures and techniques are omitted so as to not unnecessarily obscure the concepts of the present disclosure.
The terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the disclosure. The terms "comprises," "comprising," and the like, as used herein, specify the presence of stated features, steps, operations, and/or components, but do not preclude the presence or addition of one or more other features, steps, operations, or components.
All terms (including technical and scientific terms) used herein have the same meaning as commonly understood by one of ordinary skill in the art unless otherwise defined. It is noted that the terms used herein should be interpreted as having a meaning that is consistent with the context of this specification and should not be interpreted in an idealized or overly formal sense.
In those instances where a convention analogous to "at least one of A, B, and C, etc." is used, in general such a construction is intended in the sense one having skill in the art would understand the convention (e.g., "a system having at least one of A, B, and C" would include but not be limited to systems that have A alone, B alone, C alone, A and B together, A and C together, B and C together, and/or A, B, and C together, etc.).
It should be noted that the information prediction method and apparatus of the present disclosure may be used in the field of big data technology and the field of financial technology, and may also be used in any field except the field of financial technology and the field of big data technology.
In the technical scheme of the disclosure, the collection, storage, use, processing, transmission, provision, disclosure, application and other processing of the personal information of the related user are all in accordance with the regulations of related laws and regulations, necessary confidentiality measures are taken, and the customs of the public order is not violated.
At present, methods such as a neural network, an expert system, logistic regression, a decision tree and the like are widely applied, a lot of high-quality training samples need to be provided in the process of Artificial Neural Network (ANN) learning, meanwhile, the convergence rate is low in the process of training by using the artificial neural network, the obtained result is often the local best, and the like, the expert system also has the problems that knowledge is difficult to obtain, a knowledge base is inconvenient to maintain, and the like, however, the defects of the artificial neural network method can be well solved in the process of data training by using a Support Vector Machine (SVM) method, the artificial neural network method has strong processing capacity on the problems of few processing samples, nonlinearity and high dimensionality, and the like, and meanwhile, the method has strong generalization capacity and can also solve the global optimal solution, and is suitable for financial risk related data indexes in China and has the characteristics of small amount of historical data, nonlinearity, high dimensionality and the like.
However, the classification performance of the support vector machine is rather dependent on the penalty factor and the kernel function parameters. In the related art, the conventional algorithm such as grid search and the like and the group intelligent algorithm are generally used to optimize parameters of the support vector machine, for example: genetic algorithms, ant colony algorithms, and the like. However, these algorithms are easy to be trapped in local optimum, and search for a local minimum value rather than a global optimum value, so that the accuracy of the support vector machine is low when the support vector machine is applied to credit risk prediction.
In view of this, an embodiment of the present disclosure provides an information prediction method, including: responding to a credit risk prediction request for a target enterprise from a client, and acquiring operation condition information of the target enterprise from a database; inputting the operation condition information into a classification model, and outputting a classification result; the classification model is obtained by initializing sample data by utilizing chaotic mapping to obtain a sample data set and then performing iterative training on a preset model, wherein the sample data comprises the operation condition information and the credit category information of an enterprise, and the model parameter of the t +1 iterative training is determined according to the result of the t iterative training, wherein t is a positive integer; and generating credit risk prediction result information of the target enterprise according to the classification result.
Fig. 1 schematically illustrates an application scenario of information prediction according to an embodiment of the present disclosure.
As shown in fig. 1, the application scenario 100 according to this embodiment may include terminal devices 101, 102, 103, a network 104, and a server 105. Network 104 is the medium used to provide communication links between terminal devices 101, 102, 103 and server 105. Network 104 may include various connection types, such as wired, wireless communication links, or fiber optic cables, among others.
The user may use the terminal devices 101, 102, 103 to interact with the server 105 via the network 104 to receive or send messages or the like. The terminal devices 101, 102, 103 may have installed thereon various communication client applications, such as shopping-like applications, web browser applications, search-like applications, instant messaging tools, mailbox clients, social platform software, etc. (by way of example only).
The terminal devices 101, 102, 103 may be various electronic devices having a display screen and supporting web browsing, including but not limited to smart phones, tablet computers, laptop portable computers, desktop computers, and the like.
The server 105 may be a server providing various services, such as a background management server (for example only) providing support for websites browsed by users using the terminal devices 101, 102, 103. The background management server may analyze and perform other processing on the received data such as the user request, and feed back a processing result (e.g., a webpage, information, or data obtained or generated according to the user request) to the terminal device.
It should be noted that the information prediction method provided by the embodiment of the present disclosure may be generally executed by the server 105. Accordingly, the information prediction apparatus provided by the embodiment of the present disclosure may be generally disposed in the server 105. The information prediction method provided by the embodiment of the present disclosure may also be performed by a server or a server cluster that is different from the server 105 and is capable of communicating with the terminal devices 101, 102, 103 and/or the server 105. Accordingly, the information prediction apparatus provided by the embodiment of the present disclosure may also be disposed in a server or a server cluster different from the server 105 and capable of communicating with the terminal devices 101, 102, 103 and/or the server 105.
It should be understood that the number of terminal devices, networks, and servers in fig. 1 is merely illustrative. There may be any number of terminal devices, networks, and servers, as desired for an implementation.
The information prediction method of the disclosed embodiment will be described in detail below with reference to fig. 2 to 4 based on the scenario described in fig. 1.
Fig. 2 schematically shows a flow chart of an information prediction method according to an embodiment of the present disclosure.
As shown in fig. 2, the information prediction method of this embodiment includes operations S210 to S230.
In operation S210, in response to a credit risk prediction request for a target enterprise from a client, operational status information of the target enterprise is obtained from a database.
According to an embodiment of the present disclosure, the operational condition information of the target enterprise may include a total asset profit Rate (ROA), a net operating Rate (ROS), an interest support factor (ICR), an asset liability rate (DAR), a long term capital liability rate (LLR), a flow rate (LR), a snap rate (QR), a Cash Rate (CR), a total asset turnover rate (AT), an operational capital turnover rate (WCT), and the like. The above-mentioned operating condition information can be obtained from a public database, for example: CSMAR database.
In operation S220, inputting the operation condition information into the classification model, and outputting a classification result; the classification model is obtained by initializing sample data by utilizing chaotic mapping to obtain a sample data set and then performing iterative training on a preset model, wherein the sample data comprises operation condition information and credit category information of an enterprise, and model parameters of the t +1 th iterative training are determined according to a result of the t-th iterative training, wherein t is a positive integer.
According to the embodiment of the disclosure, the operation condition information can be used as an input vector and input into the classification model, and the output classification result can represent that the operation condition of the target enterprise is good or poor. For example: the output classification result close to 1 can indicate that the operation condition of the target enterprise is good. The output classification result close to-1 can indicate that the operation condition of the target enterprise is poor.
In operation S230, credit risk prediction result information of the target enterprise is generated according to the classification result.
According to an embodiment of the present disclosure, for example: the classification result is 0.9, which indicates that the operation status of the target enterprise is good, and the generated credit risk prediction result information of the target enterprise may be low in credit risk.
According to an embodiment of the present disclosure, for example: the classification result is-0.5, which indicates that the operation status of the target enterprise is poor, and the generated credit risk prediction result information of the target enterprise may be that the credit risk is high.
According to the embodiment of the disclosure, the operation condition information of the target enterprise is acquired from the database, the operation condition information is input into the classification model, the classification result is output, and the credit risk prediction result information of the target enterprise is generated according to the classification result. The classification model is obtained by chaotic mapping and initializing sample data to obtain a sample data set and then performing iterative training on the preset model, so that the optimization efficiency and the traversal uniformity of the input operation condition information are improved, and the accuracy of the credit risk prediction result is improved to a certain extent.
Fig. 3 schematically shows a flow chart of a training method of a classification model according to an embodiment of the present disclosure.
As shown in FIG. 3, the training method of the classification model includes operations S310 to S340.
In operation S310, operation status information and credit category information of n enterprises are obtained as sample data, where the operation status information of each enterprise includes m pieces of operation index information, and m and n are positive integers.
According to an embodiment of the present disclosure, the credit category information may include a category with a lower degree of credit and a category with a higher degree of credit. A less confident category of business indicates that the business has had a loan default event. A more confident category of business indicates that the business has not experienced a loan default event.
According to an embodiment of the present disclosure, each sample data may be represented as an m-dimensional vector, the credit category information may be used as a tag of the sample data, and may be represented by +1 and-1, respectively, where the "+1" tag represents that the enterprise credit degree corresponding to the sample data is high, and the "— 1" tag represents that the enterprise credit degree corresponding to the sample data is low.
In operation S320, iteration conditions and initialization parameters are determined according to application requirements of the credit risk prediction scenario.
According to an embodiment of the present disclosure, the iteration condition may include a maximum number of iterations, an iteration termination threshold, and the like. The initialization parameters may include a population scale of a particle swarm composed of sample data, a dimension of a particle, a velocity of an initial particle, a penalty parameter value range, a kernel parameter value range, and the like.
According to an embodiment of the present disclosure, the application requirements of the credit risk prediction scenario may include the prediction requirements of the credit risk without the application scenario, for example: in a loan approval application scene, the accuracy of the demand of credit risk prediction is high, the determined iteration conditions include that the maximum iteration number can be 200, and a specific iteration termination threshold value can be determined according to actual application demands.
According to an embodiment of the present disclosure, for example: in the loan approval application scenario, initialization parameters that may be set include: the population size may be 20, the maximum velocity of the initial particle may be 0.5, the minimum velocity of the initial particle may be-0.5, the penalty parameter value range may be (0, 100), and the kernel parameter value range may be (0, 100).
In operation S330, the sample data is initialized using the chaotic map according to the initialization parameter to obtain a sample data set.
According to the embodiment of the disclosure, for example, a Logistic chaotic mapping algorithm may be used to initialize sample data, each sample data is used as a particle, and the position of the particle swarm may be represented as X (c, σ), where c represents a penalty parameter and σ represents a kernel parameter, and the penalty parameter c and the kernel parameter σ are parameters to be optimized in a training process. The distribution in the solution space is random, and after chaotic mapping, the positions X (c, sigma) of the particles are uniformly distributed in the solution space in a distributed manner, so that the globally optimal particle positions can be obtained by training a classification model.
According to an embodiment of the present disclosure, for example: the chaotic mapping process can be expressed as equation (1):
CX＝(X-rand min )/(rand max -rand min )
CX′＝a×CX×(1-CX)
X′＝rand min +CX′×(rand max -rand min ) (1)
wherein CX represents a chaotic mapping variable and has a value range of [0,1%](ii) a X represents an initial particle; x' represents the particles after chaotic mapping; a represents the chaotic mapping system parameter and has the value range of 0,4]；rand min The minimum value in the value ranges of the penalty parameter and the nuclear parameter in the initialization parameter is represented; rand max And the maximum value in the value ranges of the penalty parameter and the kernel parameter in the initialization parameter is represented.
In operation S340, according to the iteration condition, the sample data set is input into the preset model for iterative training, so as to obtain a trained classification model.
According to an embodiment of the present disclosure, for example: the maximum iteration number can be set to be 200, the preset model is trained according to the maximum iteration number, the fitness of the particles in the solution space tends to converge gradually in the training process, the threshold value of the fitness can be set according to the actual application requirement, when the fitness of the particles in the solution space reaches the preset threshold value, the iterative training is completed, and the classification model after the training can be obtained.
According to the embodiment of the disclosure, the initial solution quality is improved by using Logistic chaotic mapping, the global search capability of a particle swarm algorithm is improved, the problems that the population diversity is reduced and the particle swarm algorithm is easy to fall into the local optimum when the population intelligent algorithm is close to the optimal solution are solved, the Logistic chaotic mapping is applied to the particle swarm algorithm (PSO), the distribution uniformity of the initial solution is increased, the optimization searching efficiency and the traversal uniformity are improved, the population search capability is improved, the defects that the population diversity is reduced when the population intelligent algorithm is close to the optimal solution, the particle swarm algorithm is easy to fall into the local optimum, the search precision is reduced and the like are overcome to a certain extent, and the basic particle swarm algorithm is improved and optimized, so that the global search capability and the local development capability are improved, the convergence speed of the optimization is improved, the mining capability and the algorithm efficiency of the algorithm are improved, and the classification accuracy of a classification model is improved.
According to the embodiment of the disclosure, inputting a sample data set into a preset model for iterative training to obtain a trained classification model;
inputting the sample data set into a preset model, and outputting a classification result, wherein the classification result represents the fitness of each sample data in the sample data set;
determining model parameters of the t +1 th iterative training by adopting a self-adaptive weight method according to the fitness of the sample data in the t-th iterative training;
and performing iterative training on the preset model by adopting a k-fold cross validation strategy until the classification result meets an iterative condition to obtain a trained classification model, wherein k is a positive integer which is more than 5 and less than 20.
According to an embodiment of the present disclosure, for example: in the t-th iterative training, the current position of the ith particle may be as shown in equation (2):
X i (t)＝[X i，1 (t)，X i，2 (t)，…，X i，N (t)] (2)
wherein, X i，N (t) indicates the location of the current nth operation index data for the ith enterprise.
The current velocity of the ith particle may be as shown in equation (3):
V i (t)＝[V i，1 (t)，V i，2 (t)，…，V i，N (t)] (3)
wherein, V i，N (t) represents the currentSpeed of nth operation index data of i enterprises.
The current individual optimal position of the ith particle may be as shown in equation (4):
P i (t)＝[P i，1 (t)，P i，2 (t)，…，P i，N (t)] (4)
wherein, P i，N (t) represents the current optimum location of the nth operation index data for the ith enterprise.
The process of updating the particle position in the tth iterative training can be expressed as equation (5):
Figure BDA0003826713980000121
therefore, the global optimal position of the ith particle can be as shown in equation (6):
G(t)＝P g (t)＝[P g，1 (t)，P g，2 (t)，…，P g，N (t)]，1≤g≤M (6)
the update of the particle velocity of the ith particle in the training process can be shown as formula (7), and the update of the particle position can be shown as formula (8):
V i，j (t+1)＝V i，j (t)+c 1 ·r 1，i，j (t)·(P i，j (t)-X i，j (t))+c 2 ·r 2，i，j (t)·(G j (t)-X i，j (t)) (7)
X i，j (t+1)＝V i，j (t+1)+X i，j (t) (8)
wherein i is more than or equal to 1 and less than or equal to M, and M represents the number of particle groups of the test data set obtained after chaotic mapping; j is more than or equal to 1 and less than or equal to N, and N represents the dimension of a parameter search space, namely the number of operation indexes in each enterprise in sample data; the parameter optimization of the t time when the particles are searched is represented as t; c. C 1 、c 2 A learning factor representing a population of particles, wherein c 1 A part for adjusting learning to self, depending on individual experience; c. C 2 The system is used for adjusting group learning and learning the whole particle swarm; r is 1，i，j (t) and r 2，i，j (t) is a random number generated in the range of (0, 1); to prevent the particles from jumping out of the spatial range during parameter optimization, the particle velocity range is set, i.e. v i，j (t)∈[-V max ，V max ]Similarly, the position range of the particle is set, i.e. X i，j (t)∈[-X max ，X max ]. According to the algorithm principle, the particle swarm is continuously close to the current global optimal position, and the global optimal solution is continuously updated according to the fitness, so that the particle swarm can be aggregated and concentrated at the new optimal position.
According to the embodiment of the disclosure, the training process of the classification model aims to determine the optimal classification hyperplane and the penalty parameter c and the kernel parameter σ corresponding to the decision function through an algorithm. In the process of optimizing the parameter space, in order to give consideration to global exploration and later-stage local development and improve the convergence speed of a group in the later stage, the global search capability of particles can be enhanced in the earlier stage of particle swarm search, the optimization range is expanded, and the particles are prevented from falling into local optimization as much as possible; in the later stage, the particle swarm is close to the global optimum value, the development capability of the particles to the local part can be enhanced, the neighborhood is searched, and the convergence speed is effectively accelerated.
In view of this, in the embodiment of the present disclosure, the adaptive weight method is used to determine the model parameters of the t +1 th iterative training.
According to an embodiment of the present disclosure, the adaptive weight method may be represented by equations (9), (10), (11):
Figure BDA0003826713980000131
wherein f represents the individual fitness value of the current particle in the t-1 th iterative training; f. of max Representing the individual fitness value of the optimal particle in the t-1 st iterative training; f. of avg Representing the average fitness value of the particle group in the t-1 th iterative training; w is a min Representing the minimum value of the inertia weight coefficient in the t-1 th iterative training; w is a max And representing the maximum value of the inertia weight coefficient in the t-1 th iterative training.
c 1 ＝c max -c max ×(t-1)/maxgen (10)
c 2 ＝c min +(t-1)/maxgen (11)
Where t denotes the current number of iterations, c max Represents the maximum value of the learning factor in the t-th iterative training, c min Represents the minimum value of the learning factor in the t-th iterative training. Maxgen represents the maximum number of iterations in the iteration condition.
According to the embodiment of the disclosure, the model parameter of the t +1 th iterative training can be determined in the parameter space by the adaptive weight method according to the result of the t-th iterative training. Compared with the traditional method for fixing the learning parameters and the inertia coefficients, the method has the advantages that the global search capability and the local development capability are improved, and the optimized convergence speed is improved.
According to the embodiment of the disclosure, iterative training is performed on a preset model by adopting a k-fold cross validation strategy until a classification result meets an iterative condition, so as to obtain a trained classification model, which includes:
aiming at the t-th iterative training, dividing a sample data set into k groups, taking the kth group of sample data as a test data set, and taking other k-1 groups of sample data as a training data set;
inputting k groups of sample data into a classification model obtained by training k-1 groups of sample data, and outputting a classification result;
under the condition that all sample data in the sample data set are tested, determining the classification accuracy of the t-th iterative training according to all classification results and credit category information of the t-th iterative training;
and under the condition that the classification accuracy meets the iteration condition, obtaining a trained classification model.
According to the embodiment of the disclosure, taking a ten-fold cross validation strategy with k being 10 as an example, for the t-th iterative training, the sample data set may be divided into 10 groups, the 1 st group of data is used as a test data set, and the remaining 9 groups of data are used as a training data set. Training the preset model by using the rest 9 groups of data to obtain a classification model A 1 To be connected toInput 1 set of data into the classification model A 1 In (3), obtain a classification result T 1 . The 2 nd group of data was used as the test data set, and the remaining 9 groups were used as the training data set. Training the preset model by using the rest 9 groups of data to obtain a classification model A 2 Inputting the 2 nd group of data into the classification model A 2 In order to obtain a classification result T 2 And repeating the steps until all 10 groups of data are used as the testing data set to finish the testing process, and obtaining a classification result T 1 、T 2 、...、T 10 And taking the average accuracy of the classification result as the tth iterative training as the fitness value f of the particle.
According to the embodiment of the disclosure, in the iterative training process, the fitness value is continuously converged, the highest classification accuracy can be used as the termination condition of the iteration, and at this time, a group of parameters corresponding to the particles with the highest classification accuracy are the optimal classification hyperplane and the penalty parameter c and the kernel parameter σ corresponding to the decision function, so that the classification model after the training is obtained.
According to an embodiment of the present disclosure, the decision function f (c, σ) of the classification model may be as shown in equation (12):
Figure BDA0003826713980000141
wherein x is i Representing operational status data of the target enterprise; y is i Representing the output result; k (x) i And x) represents an RBF kernel function, wherein x represents a kernel parameter sigma, and penalty parameters c, ai and b represent the optimal solution of the ith particle.
According to the embodiment of the disclosure, the classification result of each iterative training is optimized through a k-fold cross validation strategy, and the classification accuracy of the training model can be improved.
Because the enterprise operation condition data in the public database comprises operation index data with multiple dimensions, the data dimensions are higher, and higher correlation possibly exists among different types of operation index data, so that the classification accuracy of the classification model is influenced.
According to an embodiment of the present disclosure, the training method of the classification model further includes:
determining the incidence relation of the m pieces of operation index information according to the m pieces of operation index information aiming at the operation condition information of each enterprise;
determining the accumulated contribution rate of each operation index by a principal component analysis method according to the incidence relation;
and performing dimensionality reduction on the m pieces of operation index information according to the accumulated contribution rate to obtain p pieces of operation index information, wherein p is a positive integer smaller than m.
According to an embodiment of the present disclosure, for example: the operation condition information of each enterprise includes 23 different types of operation index data, and the cumulative contribution rate of the 23 different types of operation index data can be determined by a principal component analysis method, for example: the cumulative contribution rate of the top 10 principal components is 88.6%, and it can be determined that the top 10 principal component indexes substantially contain the information in all 23 operation indexes. Dimension reduction processing can be performed on 23 pieces of operation index data, and the first 10 principal component indexes are used as index types in sample data.
According to the embodiment of the disclosure, the original sample data can be subjected to dimensionality reduction through a principal component analysis method, so that the influence of overhigh dimensionality of the sample data and certain incidence relation on the accuracy rate of the classification result is effectively reduced, and the classification accuracy rate of the classification model is improved.
Generally, the credit risk prediction result of the enterprise with the classification result between [0,1] can be set as good credit, namely the credit risk is low; and setting the credit risk prediction result of the enterprise with the classification result between-1, 0 as poor credit and high credit risk. For enterprises in different operation stages, the credit risk can be predicted according to the operation trend of the enterprises.
According to an embodiment of the present disclosure, the information prediction method further includes:
determining operation trend information of the target enterprise according to the operation condition information of the target enterprise;
and generating credit risk prediction result information of the target enterprise according to the classification result and the operation trend information of the target enterprise.
According to the embodiment of the disclosure, the operation trend information of the target enterprise can be determined according to the operation condition information of the target enterprise in the preset period. For example: the operation condition of the target enterprise A in the preset period is in a growing trend, the good credit interval can be set to be-0.2,1 for the enterprises in the operation increasing stage, and the good credit interval can be set to be 0.4,1 for the enterprises in the operation decreasing stage.
According to the embodiment of the disclosure, credit risk prediction result information of the target enterprise is generated by combining the enterprise operation trend and the classification result. The influence degree of the operation index information on the credit risk under different operation stages of the enterprise is comprehensively considered, so that the credit risk prediction result is more accurate.
Based on the information prediction method, the disclosure also provides an information prediction device. The apparatus will be described in detail below with reference to fig. 4.
Fig. 4 schematically shows a block diagram of the information prediction apparatus according to an embodiment of the present disclosure.
As shown in fig. 4, the information prediction apparatus 400 of this embodiment includes an acquisition module 410, a prediction module 420, and a first generation module 430.
The obtaining module 410 is configured to obtain the operation condition information of the target enterprise from the database in response to a credit risk prediction request for the target enterprise from the client. In an embodiment, the obtaining module 410 may be configured to perform the operation S210 described above, which is not described herein again.
The prediction module 420 is configured to input the operation condition information into the classification model and output a classification result; the classification model is obtained by initializing sample data by utilizing chaotic mapping to obtain a sample data set and then performing iterative training on a preset model, wherein the sample data comprises operation condition information and credit category information of an enterprise, and model parameters of the t +1 th iterative training are determined according to a result of the t-th iterative training, wherein t is a positive integer. In an embodiment, the prediction module 420 may be configured to perform the operation S220 described above, which is not described herein again.
The first generating module 430 is configured to generate credit risk prediction result information of the target enterprise according to the classification result. In an embodiment, the first generating module 430 may be configured to perform the operation S230 described above, which is not described herein again.
According to an embodiment of the present disclosure, the prediction module includes an acquisition sub-module, a first determination sub-module, an initialization sub-module, and a training sub-module. The acquisition submodule is used for acquiring the operation condition information and credit category information of n enterprises as sample data, wherein the operation condition information of each enterprise comprises m pieces of operation index information, and m and n are positive integers. And the first determining submodule is used for determining iteration conditions and initialization parameters according to the application requirements of the credit risk prediction scene. And the initialization module is used for initializing the sample data by utilizing chaotic mapping according to the initialization parameters to obtain the sample data set. And the training submodule is used for inputting the sample data set into a preset model according to iteration conditions to carry out iteration training to obtain a trained classification model.
According to an embodiment of the present disclosure, a training submodule includes an output unit, a determination unit, and a training unit. The output unit is used for inputting the sample data set into a preset model and outputting a classification result, and the classification result represents the fitness of each sample data in the sample data set. And the determining unit is used for determining the model parameters of the t +1 th iterative training by adopting a self-adaptive weight method according to the fitness of the sample data in the t-th iterative training. And the training unit is used for carrying out iterative training on the preset model by adopting a k-fold cross validation strategy until the classification result meets the iterative condition to obtain the trained classification model, wherein k is a positive integer which is more than 5 and less than 20.
According to an embodiment of the present disclosure, a training unit includes a grouping subunit, an output subunit, a first determination subunit, and a second determination subunit. The grouping subunit is used for dividing the sample data set into k groups aiming at the t-th iterative training, taking the kth group of sample data as a test data set, and taking other k-1 groups of sample data as a training data set. And the output subunit is used for inputting the k groups of sample data into a classification model obtained by training the k-1 groups of sample data and outputting a classification result. And the first determining subunit is used for determining the classification accuracy of the t-th iterative training according to all classification results and credit category information of the t-th iterative training under the condition that all sample data in the sample data set are tested. And the second determining subunit is used for obtaining the trained classification model under the condition that the classification accuracy rate meets the iteration condition.
According to an embodiment of the present disclosure, the prediction module further includes a second determining sub-module, a third determining sub-module, and a processing sub-module. The second determining submodule is configured to determine, according to the m pieces of operation index information, an association relationship between the m pieces of operation index information for each enterprise. And the third determining submodule is used for determining the accumulated contribution rate of each operation index through a principal component analysis method according to the association relation. And the processing submodule is used for performing dimensionality reduction processing on the m pieces of operation index information according to the accumulated contribution rate to obtain p pieces of operation index information, wherein p is a positive integer smaller than m.
According to an embodiment of the present disclosure, the information prediction apparatus further includes a determining module and a second generating module. The determining module is used for determining the operation trend information of the target enterprise according to the operation condition information of the target enterprise. And the second generation module is used for generating credit risk prediction result information of the target enterprise according to the classification result and the operation trend information of the target enterprise.
According to an embodiment of the present disclosure, any plurality of the obtaining module 410, the predicting module 420 and the first generating module 430 may be combined into one module to be implemented, or any one of the modules may be split into a plurality of modules. Alternatively, at least part of the functionality of one or more of these modules may be combined with at least part of the functionality of the other modules and implemented in one module. According to an embodiment of the present disclosure, at least one of the obtaining module 410, the predicting module 420, and the first generating module 430 may be implemented at least in part as a hardware circuit, such as a Field Programmable Gate Array (FPGA), a Programmable Logic Array (PLA), a system on a chip, a system on a substrate, a system on a package, an Application Specific Integrated Circuit (ASIC), or may be implemented in hardware or firmware in any other reasonable manner of integrating or packaging a circuit, or in any one of or a suitable combination of software, hardware, and firmware. Alternatively, at least one of the obtaining module 410, the predicting module 420 and the first generating module 430 may be implemented at least partly as a computer program module, which when executed, may perform a corresponding function.
Fig. 5 schematically shows a block diagram of an electronic device adapted to implement the information prediction method according to an embodiment of the present disclosure.
As shown in fig. 5, an electronic device 500 according to an embodiment of the present disclosure includes a processor 501 that can perform various appropriate actions and processes according to a program stored in a Read Only Memory (ROM) 502 or a program loaded from a storage section 508 into a Random Access Memory (RAM) 503. The processor 501 may comprise, for example, a general purpose microprocessor (e.g., a CPU), an instruction set processor and/or associated chipset, and/or a special purpose microprocessor (e.g., an Application Specific Integrated Circuit (ASIC)), among others. The processor 501 may also include onboard memory for caching purposes. Processor 501 may include a single processing unit or multiple processing units for performing different actions of a method flow according to embodiments of the disclosure.
In the RAM 503, various programs and data necessary for the operation of the electronic apparatus 500 are stored. The processor 501, the ROM 502, and the RAM 503 are connected to each other by a bus 504. The processor 501 performs various operations of the method flows according to the embodiments of the present disclosure by executing programs in the ROM 502 and/or the RAM 503. Note that the programs may also be stored in one or more memories other than the ROM 502 and the RAM 503. The processor 501 may also perform various operations of method flows according to embodiments of the present disclosure by executing programs stored in the one or more memories.
According to an embodiment of the present disclosure, electronic device 500 may also include an input/output (I/O) interface 505, input/output (I/O) interface 505 also being connected to bus 504. The electronic device 500 may also include one or more of the following components connected to the I/O interface 505: an input portion 506 including a keyboard, a mouse, and the like; an output portion 507 including a display such as a Cathode Ray Tube (CRT), a Liquid Crystal Display (LCD), and the like, and a speaker; a storage portion 508 including a hard disk and the like; and a communication section 509 including a network interface card such as a LAN card, a modem, or the like. The communication section 509 performs communication processing via a network such as the internet. A drive 510 is also connected to the I/O interface 505 as needed. A removable medium 511 such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, or the like is mounted on the drive 510 as necessary, so that a computer program read out therefrom is mounted into the storage section 508 as necessary.
The present disclosure also provides a computer-readable storage medium, which may be embodied in the device/apparatus/system described in the above embodiments; or may exist separately and not be assembled into the device/apparatus/system. The computer-readable storage medium carries one or more programs which, when executed, implement the method according to an embodiment of the disclosure.
According to embodiments of the present disclosure, the computer-readable storage medium may be a non-volatile computer-readable storage medium, which may include, for example but is not limited to: a portable computer diskette, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), 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 present disclosure, 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. For example, according to embodiments of the present disclosure, a computer-readable storage medium may include ROM 502 and/or RAM 503 and/or one or more memories other than ROM 502 and RAM 503 described above.
Embodiments of the present disclosure also include a computer program product comprising a computer program containing program code for performing the method illustrated in the flow chart. When the computer program product runs in a computer system, the program code is used for causing the computer system to realize the item recommendation method provided by the embodiment of the disclosure.
The computer program performs the above-described functions defined in the system/apparatus of the embodiments of the present disclosure when executed by the processor 501. The systems, apparatuses, modules, units, etc. described above may be implemented by computer program modules according to embodiments of the present disclosure.
In one embodiment, the computer program may be hosted on a tangible storage medium such as an optical storage device, a magnetic storage device, or the like. In another embodiment, the computer program may also be transmitted, distributed in the form of a signal on a network medium, downloaded and installed through the communication section 509, and/or installed from the removable medium 511. The computer program containing program code may be transmitted using any suitable network medium, including but not limited to: wireless, wired, etc., or any suitable combination of the foregoing.
In such an embodiment, the computer program may be downloaded and installed from a network through the communication section 509, and/or installed from the removable medium 511. The computer program, when executed by the processor 501, performs the above-described functions defined in the system of the embodiments of the present disclosure. The systems, devices, apparatuses, modules, units, etc. described above may be implemented by computer program modules according to embodiments of the present disclosure.
In accordance with embodiments of the present disclosure, program code for executing computer programs provided by embodiments of the present disclosure may be written in any combination of one or more programming languages, and in particular, these computer programs may be implemented using high level procedural and/or object oriented programming languages, and/or assembly/machine languages. The programming language includes, but is not limited to, programming languages such as Java, C + +, python, the "C" language, or the like. The program code may execute entirely on the user computing device, partly on the user device, partly on a remote computing device, or entirely on the remote computing device or server. In the case of a remote computing device, the remote computing device may be connected to the user computing device through any kind of network, including a Local Area Network (LAN) or a Wide Area Network (WAN), or may be connected to an external computing device (e.g., through the internet using an internet service provider).
The flowchart 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 disclosure. In this regard, 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 that perform the specified functions or acts, or combinations of special purpose hardware and computer instructions.
Those skilled in the art will appreciate that various combinations and/or combinations of features recited in the various embodiments and/or claims of the present disclosure can be made, even if such combinations or combinations are not expressly recited in the present disclosure. In particular, various combinations and/or combinations of the features recited in the various embodiments and/or claims of the present disclosure may be made without departing from the spirit or teaching of the present disclosure. All such combinations and/or associations are within the scope of the present disclosure.
The embodiments of the present disclosure have been described above. However, these examples are for illustrative purposes only and are not intended to limit the scope of the present disclosure. Although the embodiments are described separately above, this does not mean that the measures in the embodiments cannot be used in advantageous combination. The scope of the disclosure is defined by the appended claims and equivalents thereof. Various alternatives and modifications can be devised by those skilled in the art without departing from the scope of the present disclosure, and such alternatives and modifications are intended to be within the scope of the present disclosure.

Claims (10)

1. An information prediction method, comprising:
responding to a credit risk prediction request for a target enterprise from a client, and acquiring operation condition information of the target enterprise from a database;
inputting the operation condition information into a classification model, and outputting a classification result; the classification model is obtained by initializing sample data by utilizing chaotic mapping to obtain a sample data set and then performing iterative training on a preset model, wherein the sample data comprises operation condition information and credit category information of an enterprise, and model parameters of the t +1 th iterative training are determined according to the result of the t-th iterative training, wherein t is a positive integer; and
and generating credit risk prediction result information of the target enterprise according to the classification result.
2. The method of claim 1, the training method of the classification model comprising:
acquiring operation condition information and credit category information of n enterprises as sample data, wherein the operation condition information of each enterprise comprises m pieces of operation index information, and m and n are positive integers;
determining iteration conditions and initialization parameters according to application requirements of a credit risk prediction scene;
initializing the sample data by utilizing chaotic mapping according to the initialization parameter to obtain a sample data set;
and inputting the sample data set into the preset model for iterative training according to the iterative conditions to obtain the trained classification model.
3. The method according to claim 2, wherein the inputting the sample data set into the preset model for iterative training results in the trained classification model;
inputting the sample data set into the preset model, and outputting a classification result, wherein the classification result represents the fitness of each sample data in the sample data set;
determining the model parameters of the t +1 th iterative training by adopting a self-adaptive weight method according to the fitness of the sample data in the t-th iterative training;
and performing iterative training on the preset model by adopting a k-fold cross validation strategy until the classification result meets the iterative condition to obtain the trained classification model, wherein k is a positive integer which is more than 5 and less than 20.
4. The method according to claim 3, wherein iteratively training the preset model by using a k-fold cross validation strategy until the classification result satisfies the iteration condition to obtain the trained classification model, and the method comprises:
aiming at the t-th iterative training, dividing the sample data set into k groups, taking the kth group of sample data as a test data set, and taking other k-1 groups of sample data as a training data set;
inputting the k groups of sample data into a classification model obtained by training the k-1 groups of sample data, and outputting a classification result;
under the condition that all sample data in the sample data set are tested, determining the classification accuracy of the t-th iterative training according to all classification results of the t-th iterative training and the credit category information;
and under the condition that the classification accuracy meets the iteration condition, obtaining the trained classification model.
5. The method of claim 2, further comprising:
determining the incidence relation of the m pieces of operation index information according to the m pieces of operation index information aiming at the operation condition information of each enterprise;
determining the accumulated contribution rate of each operation index through a principal component analysis method according to the association relation;
and performing dimensionality reduction on the m pieces of operation index information according to the accumulated contribution rate to obtain p pieces of operation index information, wherein p is a positive integer smaller than m.
6. The method of claim 1, further comprising:
determining operation trend information of a target enterprise according to operation condition information of the target enterprise;
and generating credit risk prediction result information of the target enterprise according to the classification result and the operation trend information of the target enterprise.
7. An information prediction apparatus comprising:
the acquisition module is used for responding to a credit risk prediction request aiming at a target enterprise from a client and acquiring the operation condition information of the target enterprise from a database;
the prediction module is used for inputting the operation condition information into a classification model and outputting a classification result; the classification model is obtained by initializing sample data by using chaotic mapping to obtain a sample data set and then performing iterative training on a preset model, wherein the sample data comprises operation condition information and credit category information of an enterprise, and model parameters of the t +1 th iterative training are determined according to the result of the t th iterative training, wherein t is a positive integer; and
and the first generation module is used for generating credit risk prediction result information of the target enterprise according to the classification result.
8. An electronic device, comprising:
one or more processors;
a storage device for storing one or more programs,
wherein the one or more programs, when executed by the one or more processors, cause the one or more processors to perform the method recited in any of claims 1-6.
9. A computer readable storage medium having stored thereon executable instructions which, when executed by a processor, cause the processor to perform the method according to any one of claims 1 to 6.
10. A computer program product comprising a computer program which, when executed by a processor, implements the method according to any one of claims 1 to 6.
CN202211068161.5A 2022-08-31 2022-08-31 Information prediction method, device, equipment and medium Pending CN115409635A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202211068161.5A CN115409635A (en) 2022-08-31 2022-08-31 Information prediction method, device, equipment and medium

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202211068161.5A CN115409635A (en) 2022-08-31 2022-08-31 Information prediction method, device, equipment and medium

Publications (1)

Publication Number Publication Date
CN115409635A true CN115409635A (en) 2022-11-29

Family

ID=84163172

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202211068161.5A Pending CN115409635A (en) 2022-08-31 2022-08-31 Information prediction method, device, equipment and medium

Country Status (1)

Country Link
CN (1) CN115409635A (en)

Similar Documents

Publication Publication Date Title
US10558913B1 (en) Machine-learning techniques for monotonic neural networks
US20200401894A1 (en) Machine-learning techniques for monotonic neural networks
US20210150355A1 (en) Training machine learning models using task selection policies to increase learning progress
US11403554B2 (en) Method and apparatus for providing efficient testing of systems by using artificial intelligence tools
US11315032B2 (en) Method and system for recommending content items to a user based on tensor factorization
US11900294B2 (en) Automated path-based recommendation for risk mitigation
US8190537B1 (en) Feature selection for large scale models
US11100559B2 (en) Recommendation system using linear stochastic bandits and confidence interval generation
US20210272195A1 (en) Instant Lending Decisions
US20230297847A1 (en) Machine-learning techniques for factor-level monotonic neural networks
US11521254B2 (en) Automatic tuning of machine learning parameters for non-stationary e-commerce data
US20210350175A1 (en) Key-value memory network for predicting time-series metrics of target entities
CN113743971A (en) Data processing method and device
US11651291B2 (en) Real-time predictions based on machine learning models
CN114462532A (en) Model training method, device, equipment and medium for predicting transaction risk
US20210166131A1 (en) Training spectral inference neural networks using bilevel optimization
US20190244131A1 (en) Method and system for applying machine learning approach to routing webpage traffic based on visitor attributes
CN114202417A (en) Abnormal transaction detection method, apparatus, device, medium, and program product
US20220383036A1 (en) Clustering data using neural networks based on normalized cuts
WO2023185125A1 (en) Product resource data processing method and apparatus, electronic device and storage medium
CN115759283A (en) Model interpretation method and device, electronic equipment and storage medium
CN115409635A (en) Information prediction method, device, equipment and medium
CN115063145A (en) Transaction risk factor prediction method and device, electronic equipment and storage medium
US20230195842A1 (en) Automated feature engineering for predictive modeling using deep reinforcement learning
US20220405531A1 (en) Blackbox optimization via model ensembling

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