CN118350921A - Novel agricultural operation subject credit evaluation method, device, equipment and storage medium - Google Patents
Novel agricultural operation subject credit evaluation method, device, equipment and storage medium Download PDFInfo
- Publication number
- CN118350921A CN118350921A CN202410429981.5A CN202410429981A CN118350921A CN 118350921 A CN118350921 A CN 118350921A CN 202410429981 A CN202410429981 A CN 202410429981A CN 118350921 A CN118350921 A CN 118350921A
- Authority
- CN
- China
- Prior art keywords
- novel agricultural
- credit
- main body
- data
- features
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000011156 evaluation Methods 0.000 title claims abstract description 54
- 238000003860 storage Methods 0.000 title claims abstract description 10
- 238000013210 evaluation model Methods 0.000 claims abstract description 49
- 238000000034 method Methods 0.000 claims abstract description 44
- 238000000605 extraction Methods 0.000 claims description 16
- 238000007637 random forest analysis Methods 0.000 claims description 14
- 238000003066 decision tree Methods 0.000 claims description 10
- 238000012545 processing Methods 0.000 claims description 8
- 238000004590 computer program Methods 0.000 claims description 6
- 238000012216 screening Methods 0.000 claims description 6
- 238000003672 processing method Methods 0.000 claims description 4
- 238000004458 analytical method Methods 0.000 claims description 3
- 230000007246 mechanism Effects 0.000 claims description 3
- 238000006243 chemical reaction Methods 0.000 abstract description 8
- 230000008569 process Effects 0.000 abstract description 8
- 238000012549 training Methods 0.000 abstract description 6
- 238000004891 communication Methods 0.000 description 5
- 238000004364 calculation method Methods 0.000 description 4
- 238000010586 diagram Methods 0.000 description 4
- 238000011161 development Methods 0.000 description 2
- 238000009313 farming Methods 0.000 description 2
- 238000012360 testing method Methods 0.000 description 2
- 230000002159 abnormal effect Effects 0.000 description 1
- 230000003044 adaptive effect Effects 0.000 description 1
- 238000013528 artificial neural network Methods 0.000 description 1
- 238000004422 calculation algorithm Methods 0.000 description 1
- 238000004140 cleaning Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 230000004927 fusion Effects 0.000 description 1
- 230000003862 health status Effects 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 230000008520 organization Effects 0.000 description 1
- 238000007781 pre-processing Methods 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 238000012706 support-vector machine Methods 0.000 description 1
Abstract
The invention provides a novel agricultural operation subject credit evaluation method, a device, equipment and a storage medium, wherein the novel agricultural operation subject credit evaluation method comprises the steps of obtaining novel agricultural operation subject data to be evaluated; extracting features of the novel agricultural operation main body data to be evaluated to obtain key feature data; inputting the key characteristic data into a credit evaluation model of a novel agricultural operation subject, and outputting a credit rating of the novel agricultural operation subject; the novel agricultural management main body credit evaluation model is obtained based on multi-dimensional novel agricultural management main body credit related data training acquired from various channels, and the novel agricultural management main body credit evaluation model is related to information to form novel agricultural management main body big data resources, features and sample numbers are more comprehensive, so that credit evaluation accuracy is improved, credit rating results are directly given, instead of predicting the default rate first and building the model to convert the default rate into credit grades, errors caused by a conversion process are avoided, and the novel agricultural management main body credit evaluation model is more direct and efficient.
Description
Technical Field
The invention relates to the technical field of data processing and intelligent decision making, in particular to a novel agricultural operation main body credit evaluation method and device, electronic equipment and a storage medium.
Background
In recent years, the number and scale of novel agricultural management bodies are improved year by year, the novel agricultural management bodies become middle-hardness forces for rural economic development, and the fund demand is large and financing dilemma is obvious. The credit evaluation and the establishment of the credit evaluation model are the basis for providing financing service for the novel agricultural management main body, and compared with other management main bodies, the novel agricultural management main body has the characteristics of difficult data acquisition and the like of financial data such as management capacity, development capacity and the like, and brings difficulty to the credit evaluation. The novel agricultural management main body comprises an agricultural cooperative, a family farm, a professional large household and a tap enterprise, is various in types and complex in identity, and only depends on the financial data of the main body, so that the credit condition of the main body is not objectively and accurately reflected. In addition, the existing novel agricultural operation main credit evaluation method is to calculate the credit default rate, then build a model to convert the default rate into credit grade, and the indirect evaluation mode is easy to cause data deviation in the conversion process, so that the credit evaluation result is inaccurate.
Disclosure of Invention
The invention provides a novel agricultural operation main body credit evaluation method, a novel agricultural operation main body credit evaluation device, electronic equipment and a storage medium, which are used for solving the defects that the credit evaluation result is inaccurate and only depends on financial data, so that the credit condition of a novel agricultural operation main body is objectively and accurately reflected due to the fact that data distortion is easily caused in the data conversion process of converting the default rate into the credit grade in the traditional novel agricultural operation main body credit evaluation method.
The invention provides a novel agricultural operation main body credit evaluation method, which comprises the following steps:
Acquiring novel agricultural operation subject data to be evaluated;
Performing feature extraction on the novel agricultural operation main body data to be evaluated to obtain key feature data;
Inputting the key characteristic data into a credit evaluation model of a novel agricultural operation subject, and outputting a credit rating of the novel agricultural operation subject;
the credit evaluation model of the novel agricultural management main body is trained based on multidimensional novel agricultural management main body credit related data collected from various channels.
According to the invention, the multi-dimensional novel credit evaluation method for the agricultural management main body comprises the following steps of:
Basic information of a novel agricultural management main body, main body asset and credit information are collected from an agricultural rural big data platform, wherein the asset information comprises at least one of land, facilities, factory buildings, agricultural products, agricultural machinery, subsidies and insurance;
Collecting the learning information of a novel agricultural management main principal from a learning information network;
collecting liability information, loan records, credit ratings and other information from a financial institution;
The go on an expedition letter structure collects letter information;
Collecting enterprise credit information of a novel agricultural management body from a national enterprise credit showing system;
collecting judicial information of a novel agricultural management main body from a China law enforcement information public network;
Collecting industrial and commercial information of a novel agricultural management main body from an enterprise information inquiry platform;
news information concerning agricultural rural areas is collected from platforms such as the internet.
According to the novel agricultural operation subject credit evaluation method provided by the invention, after the multidimensional novel agricultural operation subject credit related data collected from each channel, the novel agricultural operation subject credit evaluation method further comprises the following steps:
If the credit sample data of a certain dimension is less than a preset threshold value, new samples are obtained by random interpolation near the sample value by adopting an equalization processing method to synthesize new samples, and the boundary value of a few characteristic class samples is utilized, or the analysis is firstly carried out according to the few class samples which are easy to learn in the data, and then the processing is carried out on the rest few class samples which are difficult to learn, so that the generated samples are ensured to be similar to the existing samples, and the credit sample number of the dimension is expanded.
According to the method for evaluating the credit of the novel agricultural operation main body provided by the invention, the characteristic extraction is carried out on the novel agricultural operation main body data to be evaluated to obtain key characteristic data, and the method comprises the following steps:
Extracting a plurality of initial characteristics of the novel agricultural management main body data to be evaluated, and constructing an initial characteristic set based on the plurality of initial characteristics, wherein the initial characteristics comprise basic information characteristics, main body asset information characteristics, academic information characteristics, liability information characteristics, credit information characteristics, enterprise credit information characteristics, judicial information characteristics, industrial and commercial information characteristics and news information characteristics of the novel agricultural management main body;
evaluating the importance of each feature in the initial feature set and the correlation between every two features, and removing the unimportant features and the features with strong correlation from the initial feature set to obtain a key feature set;
Performing iterative computation on the key feature set by adopting a three-component search method to obtain the feature number when the prediction mean square error is minimum, and taking the feature number as the optimal number of key features;
and obtaining key feature data based on the key feature set and the optimal number of key features.
According to the novel agricultural operation main body credit evaluation method provided by the invention, the adoption of the three-part search method for carrying out iterative computation on the key feature set to obtain the feature number when the prediction mean square error is minimum comprises the following steps:
The key feature set M is evenly divided into 3 parts, and is expressed as M1, M2, M3, M= { M1, M2, M3}, and the feature numbers in M1, M2 and M3 are respectively as follows: [ m/3] +m3, [ m/3], [ m/3], m is the number of features in the key feature set;
Respectively taking { M1}, { M1, M2}, and M as searched sets, inputting corresponding sample data into a credit evaluation model of a novel agricultural operation main body based on a random forest, calculating a prediction mean square error based on an output result of the credit evaluation model of the novel agricultural operation main body based on the random forest, screening a set with the minimum prediction mean square error, and replacing M with the set if the number of samples in the set is not equal to M and the number of samples in the set is not less than a lower limit of a preset key feature number;
Repeating the steps until M is not changed or the number of samples in the set reaches the preset lower limit of the number of key features, and taking the number of features in the current set as the number of features when the prediction mean square error is minimum.
According to the novel agricultural operation subject credit evaluation method provided by the invention, the evaluation of the importance of each feature in the initial feature set and the correlation between every two features comprises the following steps:
calculating Gini coefficients of all the features in the initial feature set, and arranging all the features in the initial feature set in descending order according to the sizes of the Gini coefficients;
And calculating a correlation coefficient between every two features in the initial feature set, and using the pearson coefficient as a correlation evaluation index between feature variables, wherein the higher the absolute value of the pearson coefficient is, the greater the correlation between the two features is.
According to the novel agricultural operation subject credit evaluation method provided by the invention, the novel agricultural operation subject credit evaluation model is a random forest model, the key characteristic data is input into the novel agricultural operation subject credit evaluation model, and the novel agricultural operation subject credit rating is output, and the novel agricultural operation subject credit evaluation method comprises the following steps:
The novel agricultural operation subject credit rating R calculating method comprises the following steps:
wherein, a K-th decision tree is used, S i is the i-th sample, C k is the K-th decision tree, eta (C kSi) is the decision value of the K-th decision tree C k to the i-th sample S i, and the optimal decision value of the voting mechanism is used as the credit rating of the novel agricultural management main body.
The invention also provides a novel agricultural operation main body credit evaluation device, which comprises:
the acquisition module is used for acquiring novel agricultural operation subject data to be evaluated;
the extraction module is used for carrying out feature extraction on the novel agricultural operation main body data to be evaluated to obtain key feature data;
The output module is used for inputting the key characteristic data into a credit evaluation model of the novel agricultural management main body and outputting the credit rating of the novel agricultural management main body;
the credit evaluation model of the novel agricultural management main body is trained based on multidimensional novel agricultural management main body credit related data collected from various channels.
The invention also provides an electronic device, which comprises a memory, a processor and a computer program stored in the memory and capable of running on the processor, wherein the processor realizes the novel agricultural operation subject credit evaluation method according to any one of the above when executing the program.
The present invention also provides a non-transitory computer readable storage medium having stored thereon a computer program which, when executed by a processor, implements the novel agricultural business entity credit assessment method of any of the above.
The invention provides a novel agricultural operation subject credit evaluation method, a novel agricultural operation subject credit evaluation device, electronic equipment and a storage medium, wherein novel agricultural operation subject data to be evaluated are obtained; performing feature extraction on the novel agricultural operation main body data to be evaluated to obtain key feature data; inputting the key characteristic data into a credit evaluation model of a novel agricultural operation subject, and outputting a credit rating of the novel agricultural operation subject; the credit evaluation model of the novel agricultural management main body is obtained based on multi-dimensional novel agricultural management main body credit related data training acquired from each channel, data related to the novel agricultural management main body are collected from each industry and each channel, and the novel agricultural management main body big data resources are formed through information correlation, so that the characteristic and the sample number are more comprehensive, the credit evaluation accuracy is improved, the credit evaluation result is directly given, instead of predicting the default rate first and building the model to convert the default rate into the credit grade, errors caused by the conversion process are avoided, and the novel agricultural management main body credit evaluation model is more direct and more efficient.
Drawings
In order to more clearly illustrate the invention or the technical solutions of the prior art, the following description will briefly explain the drawings used in the embodiments or the description of the prior art, and it is obvious that the drawings in the following description are some embodiments of the invention, and other drawings can be obtained according to the drawings without inventive effort for a person skilled in the art.
FIG. 1 is a schematic flow chart of a novel agricultural operation subject credit evaluation method provided by the invention;
FIG. 2 is a second flow chart of the credit evaluation method of the novel agricultural management subject provided by the invention;
FIG. 3 is a third flow chart of the novel agricultural operation subject credit evaluation method provided by the invention;
FIG. 4 is a schematic diagram of the novel agricultural operation subject credit evaluation device;
Fig. 5 is a schematic structural diagram of an electronic device provided by the present invention.
Detailed Description
For the purpose of making the objects, technical solutions and advantages of the present invention more apparent, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings, and it is apparent that the described embodiments are some embodiments of the present invention, not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
Fig. 1 is a flowchart of a novel agricultural operation subject credit evaluation method provided by an embodiment of the present invention, and as shown in fig. 1, the novel agricultural operation subject credit evaluation method provided by the embodiment of the present invention includes:
step 101, acquiring novel agricultural operation subject data to be evaluated;
102, extracting features of the novel agricultural operation main body data to be evaluated to obtain key feature data;
Step 103, inputting the key characteristic data into a credit evaluation model of a novel agricultural operation subject, and outputting a credit rating of the novel agricultural operation subject; the credit evaluation model of the novel agricultural management main body is trained based on multidimensional novel agricultural management main body credit related data collected from various channels.
In the embodiment of the present invention, the credit evaluation model of the novel agricultural management subject is a random forest model, the inputting the key feature data into the credit evaluation model of the novel agricultural management subject, and outputting the credit rating of the novel agricultural management subject, includes:
The novel agricultural operation subject credit rating R calculating method comprises the following steps:
wherein, a K-th decision tree is used, S i is the i-th sample, C k is the K-th decision tree, eta (C kSi) is the decision value of the K-th decision tree C k to the i-th sample S i, and the optimal decision value of the voting mechanism is used as the credit rating of the novel agricultural management main body.
The credit rating results of the novel agricultural operation subject output by the embodiment of the invention are three-grade nine-grade systems, and are AAA, AA, A, BBB, BB, B, CCC, CC and C respectively, which are the same as the credit rating standards of the financial institutions on the novel agricultural operation subject.
It should be noted that the method in the embodiment of the invention is applicable to various credit evaluation models, and is not limited to machine learning algorithm credit evaluation models such as neural networks, decision trees, SVMs and the like.
In an embodiment of the present invention, the multidimensional novel agricultural operations subject credit-related data collected from various channels includes, but is not limited to:
Basic information of a novel agricultural management main body, main body asset and credit information are collected from an agricultural rural big data platform, wherein the asset information comprises at least one of land, facilities, factory buildings, agricultural products, agricultural machinery, subsidies and insurance;
Collecting the learning information of a novel agricultural management main principal from a learning information network;
collecting liability information, loan records, credit ratings and other information from a financial institution;
The go on an expedition letter structure collects letter information;
Collecting enterprise credit information of a novel agricultural management body from a national enterprise credit showing system;
collecting judicial information of a novel agricultural management main body from a China law enforcement information public network;
Collecting industrial and commercial information of a novel agricultural management main body from an enterprise information inquiry platform;
news information concerning agricultural rural areas is collected from platforms such as the internet.
In an embodiment of the present invention, the credit data includes, but is not limited to: credit rating, credit records, performance records, breach probabilities, breach loss rates, etc. The novel agricultural operation subject data collected from each channel is stored in the novel agricultural operation subject data resource library.
The traditional agricultural operation subject credit evaluation method is characterized in that financial data are acquired, the financial data are converted into loan default rate, a model is built again to convert the default rate into credit grade, and the credit evaluation result is inaccurate due to data distortion in the conversion process, and the agricultural operation subject identity is complex and only depends on the financial data, so that the credit condition of a novel agricultural operation subject is objectively and accurately reflected.
The invention provides a novel agricultural operation subject credit evaluation method, which comprises the steps of obtaining novel agricultural operation subject data to be evaluated; performing feature extraction on the novel agricultural operation main body data to be evaluated to obtain key feature data; inputting the key characteristic data into a credit evaluation model of a novel agricultural operation subject, and outputting a credit rating of the novel agricultural operation subject; the credit evaluation model of the novel agricultural management main body is obtained based on multi-dimensional novel agricultural management main body credit related data training acquired from each channel, data related to the novel agricultural management main body are collected from each industry and each channel, large data resources of the novel agricultural management main body are formed through information correlation, the characteristics and the sample number are more comprehensive, so that the credit evaluation accuracy is improved, a credit rating result is directly given, the model is built instead of the predicted default rate, the default rate is converted into the credit grade, errors caused by the conversion process are avoided, and the novel agricultural management main body credit evaluation model is more direct and efficient.
Based on any of the above embodiments, after the multi-dimensional novel agricultural operation subject credit related data collected from each channel, the method further includes:
Data preprocessing is performed on multidimensional novel agricultural operations subject credit-related data collected from various channels, including but not limited to the following:
(1) And correlating and cleaning various data related to the novel agricultural management main body, and processing missing values, abnormal values and repeated values.
(2) And (3) carrying out numerical treatment on special data, and carrying out standardization treatment on positive indexes, negative indexes and interval indexes.
(3) And (3) verifying and checking data for the data difference and inconsistency problems of the multi-platform data fusion.
(4) And (5) carrying out equalization processing on various sample data of credit rating.
In the embodiment of the invention, the equalization processing of various sample data of credit rating comprises the following steps:
If the credit sample data of a certain dimension is less than a preset threshold value, new samples are obtained by random interpolation near the sample value by adopting an equalization processing method to synthesize new samples, and the boundary value of a few characteristic class samples is utilized, or the analysis is firstly carried out according to the few class samples which are easy to learn in the data, and then the processing is carried out on the rest few class samples which are difficult to learn, so that the generated samples are ensured to be similar to the existing samples, and the credit sample number of the dimension is expanded.
In an embodiment of the present invention, the equalization processing method includes, but is not limited to, an SMOTE oversampling method, a Borderline SMOTE method, or an unbalanced data set processing (ADAPTIVE SYNTHETIC SAMPLING, ADASYN) method, to implement equalization of different types of sample data.
In an embodiment of the present invention, sample information associated with a novel agricultural entity includes, but is not limited to: principal name, organization code, principal registration address, registered funds, principal business, employee number, employee's academy, principal annual profit, land area, land circulation, facility size, agricultural machine number, planting variety, planting size, farming variety, farming size, supply chain order, annual estimated income, annual actual income, annual subsidy amount, liability amount, holding policy type, holding policy amount, holding policy value, legal name, legal gender, legal identification number, legal academy, legal health status, historical loan records, historical breach records, credit rating, breach probability, and the like.
Based on any one of the above embodiments, the feature extraction is performed on the novel agricultural operation subject data to be evaluated, to obtain key feature data, including:
Step 201, extracting a plurality of initial features of the novel agricultural management subject data to be evaluated, and constructing an initial feature set based on the plurality of initial features, wherein the initial features comprise basic information features, subject asset information features, academic information features, liability information features, credit information features, enterprise credit information features, judicial information features, industrial and commercial information features and news information features of the novel agricultural management subject;
Step 202, evaluating the importance of each feature in the initial feature set and the correlation between every two features, and eliminating unimportant features and features with strong correlation from the initial feature set to obtain a key feature set;
In the embodiment of the invention, evaluating the importance of each feature in the initial feature set and the correlation between every two features comprises the following steps:
calculating Gini coefficients of all the features in the initial feature set, and arranging all the features in the initial feature set in descending order according to the sizes of the Gini coefficients;
In the embodiment of the invention, the Gini coefficient can judge the influence degree of each feature on the credit rating, and the larger the Gini coefficient is, the more important the feature is represented, and the larger the contribution to predicting the credit rating of the novel agricultural management main body is.
The Gini coefficient calculation method of a certain characteristic comprises the following steps:
The sample data set D of a feature contains V different attributes, and the Gini coefficient of the feature attribute is calculated first: d v is a feature dataset of a certain attribute, k represents the number of types of ratings corresponding to D v, and P i represents the proportion of samples of the ith rating type to the total number of samples in D v. And then summarizing to obtain a Gini coefficient calculation formula of a certain feature D: the calculated Gini coefficient ranges from 0 to 1.
All features in the initial feature set are arranged in descending order of size of Gini coefficients.
And calculating a correlation coefficient between every two features in the initial feature set, and using the pearson coefficient as a correlation evaluation index between feature variables, wherein the higher the absolute value of the pearson coefficient is, the greater the correlation between the two features is.
In the embodiment of the invention, the calculation formula of the pearson coefficient is as follows: Cov (X, Y) is the covariance of feature X and feature Y, σx σy is the standard deviation of feature X and feature Y. The value of the pearson coefficient is calculated to be between-1 and 1, and the higher the absolute value of the pearson coefficient is, the larger the correlation between the two features is.
If the correlation is too high, the evaluation model is repeatedly calculated, and the prediction accuracy and the running speed of the evaluation model are reduced.
Thus, considering the importance and relevance of features in combination, unimportant features as well as features with strong relevance are removed from the initial feature set. For example, we delete the feature with Gini coefficient 0 in the initial feature set, find all feature pairs with absolute value of pearson coefficient larger than 0.9 from the rest features, delete the feature with relatively smaller Gini coefficient, and the rest is the first stage key feature set.
Step 203, performing iterative computation on the key feature set by adopting a three-component search method to obtain the feature number when the prediction mean square error is minimum, and taking the feature number as the optimal number of key features;
In the embodiment of the present invention, the performing iterative computation on the key feature set by using a ternary search method to obtain the feature number when the prediction mean square error is minimum includes:
The key feature set M is evenly divided into 3 parts, and is expressed as M1, M2, M3, M= { M1, M2, M3}, and the feature numbers in M1, M2 and M3 are respectively as follows: [ m/3] +m3, [ m/3], [ m/3], m is the number of features in the key feature set;
Respectively taking { M1}, { M1, M2}, and M as searched sets, inputting corresponding sample data into a credit evaluation model of a novel agricultural operation main body based on a random forest, calculating a prediction mean square error based on an output result of the credit evaluation model of the novel agricultural operation main body based on the random forest, screening a set with the minimum prediction mean square error, and replacing M with the set if the number of samples in the set is not equal to M and the number of samples in the set is not less than a lower limit of a preset key feature number;
Repeating the steps until M is not changed or the number of samples in the set reaches the preset lower limit of the number of key features, and taking the number of features in the current set as the number of features when the prediction mean square error is minimum.
In the embodiment of the invention, as the number of the features can influence the prediction accuracy of the evaluation model, different numbers of feature set samples are input into the random forest, the obtained prediction mean square deviations are different, and the feature number with high accuracy can be determined by comparing the sizes of the prediction mean square deviations.
The formula for predicting mean square error: n is the number of samples, y (x) is the true value of the sample, and y ′ (x) is the predicted value of the sample.
And (3) carrying out iterative computation on the key feature set obtained by screening in the first stage by using a three-part search method to obtain the feature number when the prediction mean square error is minimum, namely the optimal number of key features, and the feature set at the moment is the final key feature set.
The number of the searched features is M, the number of the features is M, in order to avoid the too small number of the key features, the number of the features is not less than n and is not less than 3 in advance, and as shown in fig. 2, the step of searching the optimal number of the features by the dichotomy is described as follows:
dividing M into 3 parts, and representing the parts as M1, M2, M3, M= { M1, M2, M3}, wherein the number of the features in M1, M2 and M3 is respectively as follows: [ m/3] +m%3, [ m/3], [ m/3].
Respectively taking { M1}, { M1, M2}, M as key feature sets, inputting corresponding sample data into a random forest model for training, calculating a prediction mean square error, and selecting a set with the minimum prediction mean square error, if the set is equal to M or the number of samples of the set is less than n, otherwise, replacing M with the set;
repeating the steps until M is not changed or a preset lower limit of the key feature quantity is reached, namely the current m=n, wherein the feature quantity M is the optimal quantity, and the key feature set M is the optimal feature set.
And 204, obtaining key feature data based on the key feature set and the optimal number of key features.
In order to avoid subjectivity of manual feature screening, and evaluate and screen up to hundreds of features, the dimension is reduced and the generalization capability of the model is improved.
As shown in fig. 3, the model was trained and optimized, and the data was divided into training and test sets at a ratio of 7:3. The key characteristic data is used as input, the credit rating is used as output, and the random forest is trained to obtain a credit evaluation model of the novel agricultural management main body.
And judging whether iteration is finished or not through average relative errors, further determining the optimal number and the number of leaf nodes, and finishing optimization of the random forest model.
Judging the prediction accuracy of the model by using the data of the test set, and adopting an average relative error and a calculation formula of the error:
n is the number of samples, y (x) is the true value of the sample, and y , (x) is the predicted value of the sample.
And finally, inputting the data of the novel agricultural operation subject to be evaluated into a trained random forest model to obtain the credit rating corresponding to the novel agricultural operation subject.
According to the novel agricultural operation main body credit evaluation method provided by the embodiment of the invention, data related to the novel agricultural operation main body are collected from each industry and each channel, and the large data resource of the novel agricultural operation main body is formed through information correlation, so that the characteristics and the sample number are more comprehensive; directly giving rating results of three-level nine-class credit rating standards (AAA, AA, A, BBB, BB, B, CCC, CC and C) commonly used by financial institutions, and re-establishing a model instead of predicting the default rate to convert the default rate into credit rating, so that errors caused by the conversion process are avoided, and the method is more direct and more efficient; the two-stage feature extraction method is adopted: compared with other methods which only consider the importance of the features, the method adds the screening condition of the feature relevance, adopts the three-part searching method to determine the number of the key features, and is more objective and accurate than setting by experience or guess. The key features are screened by a two-stage method, so that the prediction of the credit evaluation model of the novel agricultural management main body can be more accurate.
The novel agricultural operation subject credit evaluation device provided by the invention is described below, and the novel agricultural operation subject credit evaluation device described below and the novel agricultural operation subject credit evaluation method described above can be correspondingly referred to each other.
Fig. 4 is a schematic diagram of a novel agricultural operation main body credit evaluation device provided by the embodiment of the present invention, and as shown in fig. 4, the novel agricultural operation main body credit evaluation device provided by the embodiment of the present invention includes:
an acquisition module 401, configured to acquire new agricultural operation subject data to be evaluated;
The extracting module 402 is configured to perform feature extraction on the novel agricultural operation subject data to be evaluated, so as to obtain key feature data;
The output module 403 is configured to input the key feature data into a credit evaluation model of a novel agricultural operation subject, and output a credit rating of the novel agricultural operation subject;
the credit evaluation model of the novel agricultural management main body is trained based on multidimensional novel agricultural management main body credit related data collected from various channels.
The invention provides a novel agricultural operation subject credit evaluation device, which is characterized by acquiring novel agricultural operation subject data to be evaluated; performing feature extraction on the novel agricultural operation main body data to be evaluated to obtain key feature data; inputting the key characteristic data into a credit evaluation model of a novel agricultural operation subject, and outputting a credit rating of the novel agricultural operation subject; the credit evaluation model of the novel agricultural management main body is obtained based on multi-dimensional novel agricultural management main body credit related data training acquired from each channel, data related to the novel agricultural management main body are collected from each industry and each channel, and the novel agricultural management main body big data resources are formed through information correlation, so that the characteristic and the sample number are more comprehensive, the credit evaluation accuracy is improved, the credit evaluation result is directly given, instead of predicting the default rate first and building the model to convert the default rate into the credit grade, errors caused by the conversion process are avoided, and the novel agricultural management main body credit evaluation model is more direct and more efficient.
Fig. 5 illustrates a physical schematic diagram of an electronic device, as shown in fig. 5, which may include: processor 510, communication interface (CommunicationsInterface) 520, memory 530, and communication bus 540, wherein processor 510, communication interface 520, memory 530 complete communication with each other through communication bus 540. Processor 510 may invoke logic instructions in memory 530 to perform a novel agricultural operations subject credit assessment method comprising: acquiring novel agricultural operation subject data to be evaluated; performing feature extraction on the novel agricultural operation main body data to be evaluated to obtain key feature data; inputting the key characteristic data into a credit evaluation model of a novel agricultural operation subject, and outputting a credit rating of the novel agricultural operation subject; the credit evaluation model of the novel agricultural management main body is trained based on multidimensional novel agricultural management main body credit related data collected from various channels.
In another aspect, the present invention also provides a non-transitory computer readable storage medium having stored thereon a computer program which when executed by a processor is implemented to perform the novel agricultural business entity credit assessment method provided by the above methods, the method comprising: acquiring novel agricultural operation subject data to be evaluated; performing feature extraction on the novel agricultural operation main body data to be evaluated to obtain key feature data; inputting the key characteristic data into a credit evaluation model of a novel agricultural operation subject, and outputting a credit rating of the novel agricultural operation subject; the credit evaluation model of the novel agricultural management main body is trained based on multidimensional novel agricultural management main body credit related data collected from various channels.
The apparatus embodiments described above are merely illustrative, wherein elements illustrated as separate elements may or may not be physically separate, and elements shown as elements may or may not be physical elements, may be located in one place, or may be distributed over a plurality of network elements. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment. Those of ordinary skill in the art will understand and implement the present invention without undue burden.
From the above description of the embodiments, it will be apparent to those skilled in the art that the embodiments may be implemented by means of software plus necessary general hardware platforms, or of course may be implemented by means of hardware. Based on such understanding, the foregoing technical solutions may be embodied essentially or in part in the form of a software product, which may be stored in a computer-readable storage medium, such as a ROM/RAM, a magnetic disk, an optical disk, etc., including several instructions to cause a computer device (which may be a personal computer, a server, or a network device, etc.) to perform the various embodiments or methods of some parts of the embodiments.
Finally, it should be noted that: the above embodiments are only for illustrating the technical solution of the present invention, and are not limiting; although the invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical scheme described in the foregoing embodiments can be modified or some technical features thereof can be replaced by equivalents; such modifications and substitutions do not depart from the spirit and scope of the technical solutions of the embodiments of the present invention.
Claims (10)
1. The novel agricultural operation subject credit evaluation method is characterized by comprising the following steps:
Acquiring novel agricultural operation subject data to be evaluated;
Performing feature extraction on the novel agricultural operation main body data to be evaluated to obtain key feature data;
Inputting the key characteristic data into a credit evaluation model of a novel agricultural operation subject, and outputting a credit rating of the novel agricultural operation subject;
the credit evaluation model of the novel agricultural management main body is trained based on multidimensional novel agricultural management main body credit related data collected from various channels.
2. The method for evaluating credit of a novel agricultural management subject according to claim 1, wherein the multi-dimensional novel agricultural management subject credit-related data collected from each channel comprises:
Basic information of a novel agricultural management main body, main body asset and credit information are collected from an agricultural rural big data platform, wherein the asset information comprises at least one of land, facilities, factory buildings, agricultural products, agricultural machinery, subsidies and insurance;
Collecting the learning information of a novel agricultural management main principal from a learning information network;
collecting liability information, loan records, credit ratings and other information from a financial institution;
The go on an expedition letter structure collects letter information;
Collecting enterprise credit information of a novel agricultural management body from a national enterprise credit showing system;
collecting judicial information of a novel agricultural management main body from a China law enforcement information public network;
Collecting industrial and commercial information of a novel agricultural management main body from an enterprise information inquiry platform;
news information concerning agricultural rural areas is collected from platforms such as the internet.
3. The method for evaluating credit of a novel agricultural management subject according to claim 2, wherein after the multi-dimensional novel agricultural management subject credit-related data collected from each channel, further comprises:
If the credit sample data of a certain dimension is less than a preset threshold value, new samples are obtained by random interpolation near the sample value by adopting an equalization processing method to synthesize new samples, and the boundary value of a few characteristic class samples is utilized, or the analysis is firstly carried out according to the few class samples which are easy to learn in the data, and then the processing is carried out on the rest few class samples which are difficult to learn, so that the generated samples are ensured to be similar to the existing samples, and the credit sample number of the dimension is expanded.
4. The method for evaluating credit of a novel agricultural management subject according to claim 1, wherein the feature extraction of the novel agricultural management subject data to be evaluated to obtain key feature data comprises:
Extracting a plurality of initial characteristics of the novel agricultural management main body data to be evaluated, and constructing an initial characteristic set based on the plurality of initial characteristics, wherein the initial characteristics comprise basic information characteristics, main body asset information characteristics, academic information characteristics, liability information characteristics, credit information characteristics, enterprise credit information characteristics, judicial information characteristics, industrial and commercial information characteristics and news information characteristics of the novel agricultural management main body;
evaluating the importance of each feature in the initial feature set and the correlation between every two features, and removing the unimportant features and the features with strong correlation from the initial feature set to obtain a key feature set;
Performing iterative computation on the key feature set by adopting a three-component search method to obtain the feature number when the prediction mean square error is minimum, and taking the feature number as the optimal number of key features;
and obtaining key feature data based on the key feature set and the optimal number of key features.
5. The method for evaluating credit of a novel agricultural management subject according to claim 4, wherein the iterative computation of the key feature set by using a three-way search method to obtain feature numbers with the smallest prediction mean square error comprises:
the key feature set M is evenly divided into 3 parts, and is expressed as M1, M2, M3, M= { M1, M2, M3}, and the feature numbers in M1, M2 and M3 are respectively as follows: [ M/3] +m3, [ M/3], [ M/3], M is the number of features in the key feature set M;
Respectively taking { M1}, { M1, M2}, and M as searched sets, inputting corresponding sample data into a credit evaluation model of a novel agricultural operation main body based on a random forest, calculating a prediction mean square error based on an output result of the credit evaluation model of the novel agricultural operation main body based on the random forest, screening a set with the minimum prediction mean square error, and replacing M with the set if the number of samples in the set is not equal to M and the number of samples in the set is not less than a lower limit of a preset key feature number;
Repeating the steps until M is not changed or the number of samples in the set reaches the preset lower limit of the number of key features, and taking the number of features in the current set as the number of features when the prediction mean square error is minimum.
6. The method for evaluating credit of a novel agricultural management subject of claim 4, wherein said evaluating the importance of each feature in said initial set of features and the correlation between features comprises:
calculating Gini coefficients of all the features in the initial feature set, and arranging all the features in the initial feature set in descending order according to the sizes of the Gini coefficients;
And calculating a correlation coefficient between every two features in the initial feature set, and using the pearson coefficient as a correlation evaluation index between feature variables, wherein the higher the absolute value of the pearson coefficient is, the greater the correlation between the two features is.
7. The method for evaluating credit of a new agricultural management subject according to claim 1, wherein the credit evaluation model of the new agricultural management subject is a random forest model, the key feature data is input into the credit evaluation model of the new agricultural management subject, and the new agricultural management subject credit rating is output, comprising:
The novel agricultural operation subject credit rating R calculating method comprises the following steps:
wherein, a K-th decision tree is used, S i is the i-th sample, C k is the K-th decision tree, eta (C kSi) is the decision value of the K-th decision tree C k to the i-th sample S i, and the optimal decision value of the voting mechanism is used as the credit rating of the novel agricultural management main body.
8. Novel agricultural operation main part credit evaluation device, characterized by comprising:
the acquisition module is used for acquiring novel agricultural operation subject data to be evaluated;
the extraction module is used for carrying out feature extraction on the novel agricultural operation main body data to be evaluated to obtain key feature data;
The output module is used for inputting the key characteristic data into a credit evaluation model of the novel agricultural management main body and outputting the credit rating of the novel agricultural management main body;
the credit evaluation model of the novel agricultural management main body is trained based on multidimensional novel agricultural management main body credit related data collected from various channels.
9. An electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the processor implements the novel agricultural operations subject credit assessment method of any one of claims 1 to 7 when the program is executed by the processor.
10. A non-transitory computer readable storage medium having stored thereon a computer program, wherein the computer program when executed by a processor implements the novel agricultural business entity credit assessment method of any one of claims 1 to 7.
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN118350921A true CN118350921A (en) | 2024-07-16 |
Family
ID=
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111563706A (en) | Multivariable logistics freight volume prediction method based on LSTM network | |
| CN112132233A (en) | Criminal personnel dangerous behavior prediction method and system based on effective influence factors | |
| CN109635010B (en) | User characteristic and characteristic factor extraction and query method and system | |
| CN114048436A (en) | Construction method and construction device for forecasting enterprise financial data model | |
| CN115470962A (en) | LightGBM-based enterprise confidence loss risk prediction model construction method | |
| CN117670066B (en) | Questor management method, system, equipment and storage medium based on intelligent decision | |
| CN110634060A (en) | User credit risk assessment method, system, device and storage medium | |
| CN115357764A (en) | Abnormal data detection method and device | |
| CN115358481A (en) | Early warning and identification method, system and device for enterprise ex-situ migration | |
| CN117453764A (en) | Data mining analysis method | |
| CN116383645A (en) | Intelligent system health degree monitoring and evaluating method based on anomaly detection | |
| CN111209955A (en) | Airplane power supply system fault identification method based on deep neural network and random forest | |
| CN118350921A (en) | Novel agricultural operation subject credit evaluation method, device, equipment and storage medium | |
| CN114820074A (en) | Target user group prediction model construction method based on machine learning | |
| CN114862092A (en) | Evaluation method and device based on neural network | |
| CN115330526A (en) | Enterprise credit scoring method and device | |
| CN111654853B (en) | Data analysis method based on user information | |
| CN115239502A (en) | Analyst simulation method, analyst simulation system, electronic device and storage medium | |
| CN109829500B (en) | Position composition and automatic clustering method | |
| CN117593101B (en) | Financial risk data processing and analyzing method and system based on multidimensional data | |
| CN117745080B (en) | Multi-factor authentication-based data access control and security supervision method and system | |
| CN114358908B (en) | AI-driven collaborative enterprise multidimensional credit feature extraction and evaluation method | |
| CN115310720A (en) | Method, device and equipment for predicting use intention of old people on intelligent product | |
| CN116993139A (en) | Risk enterprise identification method, device, equipment, medium and product | |
| CN116664319A (en) | Financial policy classification system based on big data |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication |