CN112508679A - Small and micro enterprise loan risk assessment method and device and storage medium - Google Patents
Small and micro enterprise loan risk assessment method and device and storage medium Download PDFInfo
- Publication number
- CN112508679A CN112508679A CN202011278327.7A CN202011278327A CN112508679A CN 112508679 A CN112508679 A CN 112508679A CN 202011278327 A CN202011278327 A CN 202011278327A CN 112508679 A CN112508679 A CN 112508679A
- Authority
- CN
- China
- Prior art keywords
- loan
- small
- enterprise
- micro
- data
- 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
- 238000000034 method Methods 0.000 title claims abstract description 145
- 238000003860 storage Methods 0.000 title claims abstract description 11
- 238000012502 risk assessment Methods 0.000 title abstract description 11
- 230000005611 electricity Effects 0.000 claims abstract description 45
- 238000011156 evaluation Methods 0.000 claims abstract description 37
- 238000012544 monitoring process Methods 0.000 claims abstract description 27
- 238000012545 processing Methods 0.000 claims description 26
- 230000006870 function Effects 0.000 claims description 24
- 230000002159 abnormal effect Effects 0.000 claims description 12
- 238000010276 construction Methods 0.000 claims description 8
- 238000012417 linear regression Methods 0.000 claims description 7
- 238000005070 sampling Methods 0.000 claims description 5
- 238000012216 screening Methods 0.000 claims description 4
- 239000011159 matrix material Substances 0.000 description 13
- 238000004422 calculation algorithm Methods 0.000 description 6
- 238000004364 calculation method Methods 0.000 description 6
- 238000012549 training Methods 0.000 description 6
- 238000004458 analytical method Methods 0.000 description 3
- 230000006399 behavior Effects 0.000 description 3
- 239000002131 composite material Substances 0.000 description 3
- 238000010586 diagram Methods 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 3
- 230000005484 gravity Effects 0.000 description 3
- 238000011161 development Methods 0.000 description 2
- 238000004519 manufacturing process Methods 0.000 description 2
- 238000010606 normalization Methods 0.000 description 2
- 238000005457 optimization Methods 0.000 description 2
- 239000000047 product Substances 0.000 description 2
- 230000000750 progressive effect Effects 0.000 description 2
- 239000013589 supplement Substances 0.000 description 2
- 230000002194 synthesizing effect Effects 0.000 description 2
- 230000015572 biosynthetic process Effects 0.000 description 1
- 238000004590 computer program Methods 0.000 description 1
- 125000004122 cyclic group Chemical group 0.000 description 1
- 238000009826 distribution Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000011835 investigation Methods 0.000 description 1
- 238000010295 mobile communication Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 229920001296 polysiloxane Polymers 0.000 description 1
- 238000007781 pre-processing Methods 0.000 description 1
- 238000003672 processing method Methods 0.000 description 1
- 230000001737 promoting effect Effects 0.000 description 1
- 238000012797 qualification Methods 0.000 description 1
- 238000013077 scoring method Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 238000003786 synthesis reaction Methods 0.000 description 1
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
- G06Q40/03—Credit; Loans; Processing thereof
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 application discloses a small and micro enterprise loan risk assessment method, a device and a storage medium, which relate to the technical field of credit approval and solve the problems that in the prior art, the repayment capability of a small and micro enterprise is inaccurately assessed and a large amount of labor is occupied; the method comprises the following steps: acquiring power utilization running state data and loan data of a small and micro enterprise; constructing a loan assessment index system of a small and micro enterprise; giving weight to each evaluation index in a loan evaluation index system, and determining an enterprise comprehensive score which is equal to the weighted sum of a plurality of evaluation indexes; setting a monitoring threshold value according to the enterprise comprehensive score to monitor the pre-loan risk and the post-loan risk of the small and micro enterprises; the method effectively solves the problems that the repayment capability of workers is inaccurate and a large amount of labor is occupied, further realizes the loan risk determination of small and micro enterprises according to the electricity utilization running state data and other data, improves the accuracy of risk assessment, and saves a large amount of personnel cost.
Description
Technical Field
The application relates to the technical field of credit approval, in particular to a small and micro enterprise loan risk assessment method and device and a storage medium.
Background
The loan is the mode that the financial institution provides the fund for the consumer, the consumer returns the fund according to a certain interest rate and return conditions, and the financial institution puts in the money in the loan mode, thereby meeting the demand of social expanded production on the fund supplement and promoting the economic development. Applying for loan also becomes the consumption habit of most of the current consumers.
At present, along with the development of society, small and micro enterprises are more and more, the small and micro enterprises also become the main force for solving the employment rate, the small and micro enterprises also become the main group with loan, and for financial institutions, how to evaluate the repayment capacity of the small and micro enterprises is the primary problem.
The existing assessment method is based on the investigation of the qualification, credit and property conditions of the small and micro enterprises by investigators, namely, the loan data of the small and micro enterprises are examined manually, so that the assessment of repayment capability of workers is inaccurate, and a large amount of labor is occupied.
Disclosure of Invention
The loan assessment method, the loan assessment device and the loan assessment storage medium for the small and micro enterprises solve the problems that in the prior art, the repayment capacity of the small and micro enterprises is not accurately assessed, and a large amount of labor is occupied.
In a first aspect, an embodiment of the present invention provides a small-micro enterprise loan assessment method, which is characterized by including:
acquiring power utilization running state data and loan data of a small and micro enterprise;
constructing a loan assessment index system of the small and micro enterprise;
weighting each assessment index in the loan assessment index system, and determining an enterprise comprehensive score which is equal to the weighted sum of the assessment indexes;
and setting a monitoring threshold value according to the enterprise comprehensive score so as to monitor the pre-loan risk and the post-loan risk of the small and micro enterprise.
With reference to the first aspect, in a possible implementation manner, the weighting of each assessment index in the loan assessment index system includes:
and determining the weight of each evaluation index by adopting an analytic hierarchy process or an entropy method.
With reference to the first aspect, in a possible implementation manner, the weighting of each assessment index in the loan assessment index system includes:
determining a weight of each of the evaluation indicators using a comprehensive weighting method, the function of the comprehensive weighting method being:
wherein, WjWeight value for comprehensive empowerment, W1jTo determine the weight value, W, by using an analytic hierarchy process and a TOPSIS process2jThe weight values are determined for the entropy method and the TOPSIS method.
With reference to the first aspect, in a possible implementation manner, the setting a monitoring threshold according to the enterprise composite score includes:
and predicting the monthly power consumption of the small and micro enterprises by adopting a gradient lifting tree method, and further setting the monitoring threshold.
With reference to the first aspect, in a possible implementation manner, the method further includes:
associating the electricity utilization running state data with the loan data, and screening fields to form a data width table;
and processing the data on the basis of the data wide table so as to solve the problem of sample imbalance, process abnormal values and fill up missing values.
With reference to the first aspect, in a possible implementation manner, the processing data on the basis of the data wide table includes:
an undersampling method or a weight-adjusting sampling method is adopted to solve the problem of sample imbalance;
processing abnormal values by adopting a box line method;
and filling up missing values by adopting a linear regression method.
With reference to the first aspect, in a possible implementation manner, the scoring system includes: payment credit, electricity utilization stability, industry landscape, electricity utilization increase length and default electricity utilization information.
In a second aspect, an embodiment of the present invention provides a small-micro enterprise loan assessment apparatus, including:
the data acquisition unit is used for acquiring power utilization running state data and loan data of the small and micro enterprise;
the system construction unit is used for constructing a loan assessment index system of the small and micro enterprise;
the comprehensive scoring unit is used for giving weight to each assessment index in the loan assessment index system and determining an enterprise comprehensive score, and the enterprise comprehensive score is equal to the weighted sum of the assessment indexes;
and the monitoring scheme generating unit is used for setting a monitoring threshold according to the enterprise comprehensive score so as to monitor the pre-loan risk and the post-loan risk of the small and micro enterprise.
With reference to the second aspect, in a possible implementation manner, the comprehensive scoring unit is specifically configured to:
and determining the weight of each evaluation index by adopting an analytic hierarchy process and/or an entropy method, and then determining the enterprise comprehensive score.
With reference to the second aspect, in a possible implementation manner, the comprehensive scoring unit is specifically configured to:
determining the weight of each evaluation index by adopting a comprehensive weighting method, and then determining an enterprise comprehensive score, wherein the function of the comprehensive weighting method is as follows:
wherein, WjWeight value for comprehensive empowerment, W1jDetermination of weight values, W, for analytic hierarchy Process and TOPSIS Process2jThe weight values are determined for the entropy method and the TOPSIS method.
With reference to the second aspect, in a possible implementation manner, the monitoring scheme generating unit is specifically configured to: and predicting the monthly power consumption of the small and micro enterprises by adopting a gradient lifting tree method, and further setting the monitoring threshold.
With reference to the second aspect, in a possible implementation manner, the apparatus further includes a data wide table constructing unit and a data processing unit:
the data width table construction unit is used for associating the electricity utilization running state data with the loan data and screening fields to form a data width table;
the data processing unit is used for processing data on the basis of the data wide table so as to solve the problem of sample imbalance, process abnormal values and fill missing values.
With reference to the second aspect, in a possible implementation manner, the data processing unit is specifically configured to:
an undersampling method or a weight-adjusting sampling method is adopted to solve the problem of sample imbalance;
processing abnormal values by adopting a box line method;
and filling up missing values by adopting a linear regression method.
With reference to the second aspect, in one possible implementation manner, the loan assessment index system includes: payment credit, electricity utilization stability, industry landscape, electricity utilization increase length and default electricity utilization information.
In a third aspect, embodiments of the present application provide a hierarchical analysis based scoring apparatus, which includes a memory and a processor;
the memory is to store computer-executable instructions;
the processor is configured to execute the computer-executable instructions to implement the method of the first aspect as well as various possible implementations of the first aspect.
In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium, where the computer-readable storage medium stores executable instructions, and when the computer executes the executable instructions, the computer is capable of implementing the method according to the first aspect and various possible implementation manners of the first aspect.
The technical scheme provided by the embodiment of the invention at least has the following technical effects or advantages:
the embodiment of the invention provides a loan risk assessment method for small and micro enterprises, which is characterized in that a loan assessment index system of the small and micro enterprises is constructed by using electricity utilization running state data and loan data, an enterprise comprehensive score is determined after each assessment index in the loan assessment index system is given weight, and pre-loan risks and post-loan risks of the small and micro enterprises can be monitored after a monitoring threshold value is set through the enterprise comprehensive score, so that the problems that workers cannot accurately assess repayment capacity and occupy a large amount of labor force are effectively solved, the loan risk assessment of the small and micro enterprises is determined according to the electricity utilization running state data and other data, the accuracy of risk assessment is improved, and a large amount of personnel cost is saved.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings needed to be used in the description of the embodiments of the present invention will be briefly introduced below, and it is obvious that the drawings in the following description are some embodiments of the present invention, and it is obvious for those skilled in the art to obtain other drawings based on these drawings without creative efforts.
Fig. 1 is a flowchart of a small-micro enterprise loan assessment method according to an embodiment of the present disclosure;
fig. 2 is a flowchart of a hierarchical analysis method of a small-micro enterprise loan assessment method according to an embodiment of the present application;
fig. 3 is a flowchart of an entropy method of a small-micro enterprise loan assessment method according to an embodiment of the present application;
FIG. 4 is a flowchart of the TIPSIS method of the small micro-enterprise loan assessment method provided in the embodiment of the present application;
fig. 5 is a flowchart of a comprehensive scoring method of a small and micro enterprise loan assessment method according to an embodiment of the application;
fig. 6 is a line diagram of a monthly power consumption prediction performed by the XGBoost algorithm of the small and micro enterprise loan assessment method according to the embodiment of the present application;
fig. 7 is a flowchart of specific processing data of a small-micro enterprise loan assessment method according to an embodiment of the present application;
fig. 8 is a schematic view of an evaluation system of a small-business loan evaluation method according to an embodiment of the present application;
fig. 9 is a schematic view of a small-micro corporation loan assessment apparatus provided in an embodiment of the present application;
fig. 10 is a schematic diagram of a small-enterprise loan assessment entity apparatus according to an embodiment of the present application.
Detailed Description
The technical solution in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. It is to be understood that the embodiments described are only a few embodiments of the present invention, and not all embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
The application provides a small-micro enterprise loan assessment method, as shown in fig. 1, the method includes the following steps S101 to S104, and the specific steps are as follows.
And S101, acquiring power utilization running state data and loan data of the small and micro enterprise.
And S102, constructing a loan assessment index system of the small and micro enterprise.
And step S103, giving weight to each evaluation index in the loan evaluation index system, and determining an enterprise comprehensive score, wherein the enterprise comprehensive score is equal to the weighted sum of the evaluation indexes.
And step S104, setting a monitoring threshold according to the enterprise comprehensive score so as to monitor the pre-loan risk and the post-loan risk of the small and micro enterprises.
According to the loan risk assessment method for the small and micro enterprises, the loan assessment index system of the small and micro enterprises is constructed by using the electricity utilization running state data and the loan data, the enterprise comprehensive score is determined after each assessment index in the loan assessment index system is given weight, pre-loan risks and post-loan risks of the small and micro enterprises can be monitored after the monitoring threshold is set through the enterprise comprehensive score, the problems that workers cannot accurately assess repayment capacity and occupy a large amount of labor force are effectively solved, the loan risk of the small and micro enterprises is determined according to the electricity utilization running state data and other data, the loan risk assessment accuracy of the small and micro enterprises is improved, and a large amount of personnel cost is saved.
Specifically, when the step S101 is performed, the power consumption operation state data of the small and micro enterprise in recent years is obtained, statistics is performed on the power consumption behavior, the payment behavior, and the default behavior to obtain relevant indexes, and the related indexes cooperate with financial institutions to obtain loan data of the small and micro enterprise in recent years.
In step S102, a loan assessment index system of the small and micro enterprise is constructed according to the electric operating state data and the loan data obtained in step S101, and specifically, the loan assessment index system is constructed according to the electricity consumption increase length, the industry perspective, the electricity consumption stability, the payment credit degree, the default electricity consumption and the loan increase width in the electric operating state data.
In step S103, specifically, the weighting of each evaluation index in the loan evaluation index system includes:
and determining the weight of each evaluation index by adopting an analytic hierarchy process or an entropy method.
In the above-described determination of the weight of each evaluation index by using the analytic hierarchy process, as shown in fig. 2, the following steps are specifically included.
Step S201, a hierarchical progressive structure model is established.
Step S202, constructing all judgment matrixes in each layer.
And step S203, sorting the hierarchical list and checking consistency.
And step S204, carrying out overall hierarchical ordering and consistency check.
When the analytic hierarchy process is specifically used, a judgment matrix of the analytic hierarchy process is obtained, 5 important indexes of credit evaluation of small micro-enterprises are constructed, and the analytic hierarchy process judgment matrix shown in the following table is obtained. The following table is referred to as table i in the following description.
In Table IThe descriptions are as follows: 1 indicates that the two elements have the same importance compared; 3 indicates that the former is slightly more important than the latter in comparison with the two elements; 5 indicates that the former is significantly more important than the latter. The reciprocal indicates if the ratio of the importance of element i to element j is aijThen the ratio of the importance of element j to element i is
And solving each hierarchy judgment matrix to obtain the weight of the analytic hierarchy process. And obtaining the weight value of each evaluation index determined by the analytic hierarchy process through the calculation of the steps.
Specifically, when the entropy method is used, as shown in fig. 3, the method includes the following steps of calculating a weight of each evaluation index determined by the entropy method.
Step S301, a training sample is constructed.
Step S302, a training sample is normalized.
Step S303, the specific gravity value of the index values of the ith item and the jth item is calculated.
Step S304, index information entropy is calculated.
In step S305, information entropy redundancy is calculated.
Step S306, index weight is calculated.
The training samples constructed in step S301 above are:
x{{x11,x12,x13...x1j},{x21,x22,x23...x2j},{x31,x32,x33...x3j}...{xi1,xi2,x3i...xij}}。
specifically, after the training samples constructed in step S301 are normalized, the following results are obtained: the forward direction index is as follows:
negative direction index:
wherein xi'jRepresenting the normalized data value; min { xjRepresents the minimum value of a group of numbers; max { xjDenotes the maximum value of a set of elements.
Further, step S303 is implemented: and calculating the specific gravity values of the i-th term index value and the j-th term index value. The calculation formula is as follows:
wherein, yijThe specific gravity value of the ith index value and the jth index value is shown, and m represents the number of tuples. Then, step S304 is performed: and calculating index information entropy. The calculation formula is as follows:
wherein ejIndicating the entropy of the index information, k is expressed as
The formula for calculating the information entropy redundancy in step S305 is: dj＝1-ej. Wherein d isjRepresenting the information entropy redundancy.
The formula for calculating the index weight in step S306 is:
wherein, WiA weight value representing each evaluation index, and n represents n elements.
Through the calculation of the steps, the index weight value determined by the entropy method can be obtained.
Further, the TOPSIS method (Technique for Order Preference by Similarity to an Ideal Solution distance) can be used to determine the weight of each evaluation index, as shown in FIG. 4, which includes the following steps.
Step S401, unifying monotonicity of each evaluation index.
Step S402, normalization processing.
In step S403, weighting processing is performed.
Step S404, determining an optimal scheme and a worst scheme.
Step S405, calculating the distance between each evaluation object and the optimal and worst scheme.
In step S406, a comprehensive evaluation value is calculated.
In step S401, in determining monotonicity of the same index of each item, the formula is used:
wherein x isijRepresenting the elements of i rows and j columns in the matrix.
In step S402, the normalization process is performed according to the formula
Wherein xijElements in the matrix denoted i rows and j columns, minjMinimum element, max, representing column jjRepresenting the largest element of the j column.
Further, the normalized data is weighted, and in step S403, the formula of the weighting process is: zij＝Wij*AijWherein Z isijScore matrix, W, representing weightsijRepresenting manually given weight values, AijRepresenting the normalized evaluation matrix.
In step S404, an optimal solution and a worst solution are determined, and the formula for determining the optimal solution is:
the formula for determining the worst case is:
wherein Z is+Representation matrix ZijMaximum solution, Z, calculated by row-Representation matrix ZijMinimum solutions calculated by row.
In step S405, the distances from each evaluation object to the optimal and worst case are calculated, respectively
And
the specific calculation formula is as follows:
wherein, XiA matrix of objects is represented that is,
the object is represented as the optimal solution distance,
representing the object as the worst case distance, Z+Representing the optimal solution matrix, Z-The worst case matrix is represented.
In step S406, a comprehensive evaluation value is calculated. The calculation formula for calculating the comprehensive evaluation value is as follows:
wherein, BiA comprehensive evaluation value matrix is represented,
the distance at which the object is the optimal solution is represented,
representing the object as the best solution distance.
Further, in step S103, the weighting of each evaluation index in the loan evaluation index system includes:
determining the weight of each evaluation index by adopting a comprehensive weighting method, wherein the function of the comprehensive weighting method is as follows:
wherein, WjWeight values, W, representing integrated weightings1jRepresents the weight values, W, determined by the analytic hierarchy process and TOPSIS process2jMethod for representing entropy values andweight values determined by the TOPSIS method.
The weighted values obtained by the analytic hierarchy process are substituted into the TOPSIS process to obtain the optimal weighted value by the subjective experience process, and the optimal weighted value is shown in the following table, which is called as table II below.
| Serial number | Index (I) | Weight of |
| 1 | Degree of payment | 0.33 |
| 2 | Stability of electricity utilization | 0.06 |
| 3 | Degree of industrial landscape | 0.21 |
| 4 | Length increased by electricity | 0.10 |
| 5 | Default power consumption | 0.30 |
Because the optimal weight value calculated according to the subjective experience method cannot be persuasive well, the importance of the index is more inclined to personal preference, and the deviation of the index weight from the actual weight exists. Therefore, an entropy method is introduced, and the weight distribution of the index is divided according to the information quantity contained in the data. The entropy method and the TOPSIS method are used for comprehensively calculating the objective optimal weight, and the objective weight is shown in the following table which is called as table III.
| Serial number | Index (I) | Weight of |
| 1 | Degree of payment | 0.35 |
| 2 | Stability of electricity utilization | 0.14 |
| 3 | Degree of industrial landscape | 0.32 |
| 4 | Length increased by electricity | 0.08 |
| 5 | Default power consumption | 0.11 |
Heddle in situIn the formula of the combined empowerment, W1jRepresents the weight values, W, determined by the analytic hierarchy process and TOPSIS process2jRepresenting the weight values determined by the entropy method and the TOPSIS method. W1jAnd W2jAnd substituting the values into a function formula of the comprehensive weighting method to obtain the final weight value according to the table II and the table III. As shown in the following Table, the following Table is referred to as Table IV.
| Serial number | Index (I) | Weight of |
| 1 | Degree of payment | 0.34 |
| 2 | Stability of electricity utilization | 0.1 |
| 3 | Degree of industrial landscape | 0.265 |
| 4 | Length increased by electricity | 0.09 |
| 5 | Default power consumption | 0.205 |
The weights found in table iv are the final weights for the several indicators illustrated.
In step S103, a composite score of the small business is obtained, and the credit score is specifically performed on the small business according to the final weight value in the table iv, and since dimensions of each index are not consistent, the obtained data needs to be further normalized to eliminate the dimensions. The formula for eliminating the dimension is:
the above formula yields the business' credit score, which results in the following table, collectively referred to as table v below.
Table V shows the pre-mortgage risk assessment scores for the business based on the steps and algorithms described above.
In step S104, a monitoring threshold is set according to the enterprise composite score, as shown in fig. 5, the method includes: and predicting the monthly power consumption of the small and micro enterprises by adopting a gradient lifting tree method, and further setting a monitoring threshold value.
Furthermore, the method for predicting the monthly power consumption of the small and micro enterprises by adopting the gradient lifting tree method can be divided into the following steps.
Step S501, a regression tree is defined, feature splitting is continuously carried out to grow a tree, a tree is added every time, a new function is learned, and the residual error predicted last time is fitted.
Step S502, an objective function of the XGboost is defined, wherein the XGboost is a gradient boost algorithm.
And step S503, fitting the residual error predicted last time by using the newly generated tree to obtain the objective function of the prediction score.
Step S504, a first derivative and a second derivative are obtained for the objective function: an optimal prediction score for each leaf node is obtained.
And step S505, the optimal prediction score is brought into the objective function, and the minimum loss is solved.
In step S506, the best branch is found through a greedy algorithm.
In step S501, after training is completed, K trees are obtained, and a sample score is predicted:
where K denotes the total number of trees, K denotes the kth tree, fx(xi) Representing the regression value at this leaf node on a tree.
In step S502, the target function of XGBoost is:
wherein,
representing the gap used to measure the prediction score and the truth score,
the regularization term is represented, T represents the number of leaf nodes, w represents the fraction of the leaf nodes, gamma represents the number of control leaf nodes, and lambda represents that the fraction of the control leaf nodes is not too large.
In step S503, the objective function of the prediction score is:
wherein,
denotes the difference between the measured prediction score and the true score, Ω (f)k) Representing a regularization term, fx(xi) Representing the regression value at this leaf node on a tree.
In step S504, the optimal prediction score for each leaf node is:
wherein, giAnd hiRepresents the removal constants:
in step S505, the minimum loss obtained by the solution is:
wherein obj represents the node score of the tree, T represents the number of leaf nodes, gamma represents the number of control leaf nodes, lambda represents that the score of the control leaf nodes is not too large, and giAnd hiRespectively, remove constants:
in step S506, an optimal branch is found, and the difference of the structure scores after the branch is:
wherein,
the score of the left sub-tree is represented,
the score of the right sub-tree is represented,
representing indivisible divisions we can reachThe number, γ, represents the complexity cost introduced by adding a new leaf node.
According to the electricity utilization running state data of the small and micro enterprises in recent years, the XGboost algorithm is used for calculating the pre-credit and post-credit enterprise index monitoring threshold values. For example, in fig. 6, the XGBoost algorithm is used to predict the power consumption of the small and micro enterprise in 7 months of 2020, and the power consumption is compared with the power consumption in a cyclic manner to set the fluctuation threshold of the power consumption.
When the prediction is further carried out, the same ratio and the ring ratio of the monthly electricity consumption of the small and micro enterprises are used as reference standards of monitoring indexes, yellow early warning is defined when the fluctuation amount of any one index of the same ratio and the ring ratio to any direction is greater than or equal to 5% and less than or equal to 10%, yellow early warning is defined when the fluctuation amount to any direction is greater than 10%, and early warning is not generated when the fluctuation amount to any direction is less than 5%.
For example, the following table is a monitoring situation of part of the enterprise 2020 in July, and is referred to as table XI below.
| Name of an enterprise | The related industries | Monitoring month | Rating of evaluation |
| Enterprise 1 | Manufacturing industry | Year 2020, month 7 | Yellow colour |
| Enterprise 2 | Construction industry | Year 2020, month 7 | Is normal |
| Enterprise 3 | Agriculture and forestry | Year 2020, month 7 | Red colour |
The loan assessment method for small and micro enterprises provided by the embodiment of the invention, as shown in fig. 7, further includes step S701 and step S702.
Step S701, the electricity utilization operation state data and the loan data are related, and fields are screened to form a data width table.
Step S702, processing the data on the basis of the data wide table so as to solve the sample imbalance phenomenon, process abnormal values and fill up missing values.
In step S701, fields that are likely to be used are screened out from the existing electricity consumption operation state data and loan data of the small and micro enterprise to form a data wide table, so as to lay a foundation for preprocessing the data on the basis of the wide table.
Further, in step S702, the data is processed on the basis of the data width table, as shown in fig. 8, including steps S801 to S803.
Step S801, an undersampling method or a weight-adjusted sampling method is adopted to solve the problem of sample imbalance.
And step S802, processing the abnormal value by adopting a box line method.
And step S803, filling the missing value by adopting a linear regression method.
In step S801, the sample data often has an unbalanced phenomenon, and the processing method for this phenomenon is generally divided into: undersampling, adjusting weights and synthesizing a few classes of oversampling techniques, abbreviated SMOTE. SMOTE is a method for processing data sample imbalance, and compared with a random oversampling processing technology, SMOTE has the advantages that a small number of types of samples can be used for analysis, manual synthesis can be performed according to the small number of samples, new samples are generated and expanded into data samples, and the processing adopts a technology of synthesizing a small number of oversampling technologies to perform data imbalance processing.
In step S802, the abnormal value is processed, and the box line method employed in the present embodiment is used to process the abnormal value. Specifically, a box diagram is drawn for the power consumption of the small and micro enterprise, data with a quantile larger than 99% are highlighted, and data with a quantile smaller than 1% are removed.
The linear regression in step S803 is one of the most common methods for solving the regression problem, and is actually a linear modeling method, and can be solved by a convex optimization method, specifically, by minimizing the following objective function:
wherein J (theta) is WJA function of X and y. The objective function comprises two parts of contents, the regular term is used for controlling the complexity of the model, the loss term is used for measuring the fitting error, and the objective function is a convex function of the same cabin W. Regularization term parameter λ>0 provides a compromise between minimizing error and model complexity to avoid over-fitting.
The linear regression method may employ the following steps.
The method comprises the following steps: given a set of training data samples D { (x)1,y1),(x2,y2),...,(xm,ym)}，yiE to R, selecting an initial value theta0And giving a convergence tolerance epsilon and a maximum iteration number K, and then solving the following optimization problem.
Step two: update theta by adopting the following formula
And when the | J (theta (K +1)) -J (theta (K)) | < epsilon or when K is equal to K, outputting 0, wherein K is the iteration number, and otherwise, repeating the step two until the condition is met.
Constructing a regression decision function: f (x) θTAnd X, performing missing value supplement by using a regression decision function to obtain complete data without data value missing so as to score the small and micro enterprise in the subsequent process.
The scoring system provided by the present application is shown in fig. 8, and includes: payment credit, electricity consumption stability, industry landscape, electricity consumption length increase, default electricity consumption information, and loan spread.
The payment credit degree comprises the following steps: the pre-stored amount of the payment, the timely rate of the payment, the amount of the arrearage and the average length of the arrearage.
The electricity utilization stability comprises: the voltage stability and the power consumption of the enterprise increase rapidly.
The industrial landscape includes: industry prosperity index and industry total electricity consumption.
The electricity utilization growth degree comprises: the monthly power consumption of the enterprises is compared with each other, and the monthly power consumption of the enterprises is compared circularly.
The default electricity utilization comprises: the number of times of illegal electricity utilization, the amount of illegal electricity stealing, the number of times of illegal electricity stealing and the amount of supplemented electricity.
The loan spread comprises: loan spread index and loan amount.
And constructing a scoring system according to the data items, and evaluating the loan risk of the small and micro enterprises.
Although the present application provides method steps as described in the text or flowcharts above, additional or fewer steps may be included based on conventional or non-inventive efforts. The sequence of steps recited in this embodiment is only one of many steps performed and does not represent a unique order of execution. When an actual apparatus or client product executes, it can execute sequentially or in parallel (e.g., in the context of parallel processors or multi-threaded processing) according to the methods shown in this embodiment or the figures.
The embodiment of the invention also provides a small and micro enterprise loan assessment device, which comprises: the system comprises a data acquisition unit 901, a system construction unit 902, a comprehensive scoring unit 903 and a monitoring scheme generation unit 904.
The data acquisition unit 901 is used for acquiring power utilization running state data and loan data of a small and micro enterprise; the system construction unit 902 is used for constructing a loan assessment index system of the small and micro enterprise; the comprehensive scoring unit 903 is used for giving weight to each assessment index in the loan assessment index system and determining enterprise comprehensive scoring, wherein the enterprise comprehensive scoring is equal to the weighted sum of the assessment indexes; the monitoring scheme generating unit 904 is configured to set a monitoring threshold according to the enterprise comprehensive score, so as to monitor pre-loan and post-loan risks of the small and micro enterprises.
The comprehensive scoring unit may be specifically configured to: and determining the weight of each evaluation index by adopting an analytic hierarchy process and/or an entropy method, and then determining the enterprise comprehensive score.
The comprehensive scoring unit may be further specifically configured to: determining the weight of each evaluation index by adopting a comprehensive weighting method, and then determining the enterprise comprehensive score, wherein the function of the comprehensive weighting method is as follows:
wherein, WjWeight value for comprehensive empowerment, W1jDetermination of weight values, W, for analytic hierarchy Process and TOPSIS Process2jThe weight values are determined for the entropy method and the TOPSIS method.
The monitoring scheme generation unit is specifically configured to: and predicting the monthly power consumption of the small and micro enterprises by adopting a gradient lifting tree method, and further setting a monitoring threshold value.
The device also comprises a data wide table construction unit and a data processing unit: the data width table construction unit is used for associating the electricity utilization operation state data with the loan data and screening fields to form a data width table; the data processing unit is used for processing the data on the basis of the data wide table so as to solve the problem of sample imbalance, process abnormal values and fill up missing values.
The data processing unit is specifically configured to: an undersampling method or a weight-adjusting sampling method is adopted to solve the problem of sample imbalance; processing abnormal values by adopting a box line method; and filling up missing values by adopting a linear regression method.
The loan assessment index system comprises: payment credit, electricity utilization stability, industry landscape, electricity utilization increase length and default electricity utilization information.
The apparatus or module according to the embodiments of the present invention may be implemented by a computer chip or an entity, or implemented by a product with certain functions. For convenience of description, the above devices are described as being divided into various modules by functions, and are described separately. The functionality of the modules may be implemented in the same one or more software and/or hardware implementations of the present application. Of course, a module that implements a certain function may be implemented by a plurality of sub-modules or sub-units in combination.
As shown in fig. 10, an embodiment of the present invention further provides a small-micro enterprise loan assessment apparatus, including: a memory 1001 and a processor 1002; memory 1001 is used to store computer executable instructions; the processor 1002 is configured to execute computer-executable instructions to implement the small micro-enterprise loan assessment method provided by the present embodiment.
The embodiment of the invention also provides a computer-readable storage medium, wherein the computer-readable storage medium stores executable instructions, and when the computer executes the executable instructions, the method for evaluating the small and micro enterprise loan provided by the embodiment is realized.
The storage medium includes, but is not limited to, a Random Access Memory (RAM), a Read-Only Memory (ROM), a Cache, a Hard Disk (Hard Disk Drive), or a Memory Card (HDD). The memory may be used to store computer program instructions.
The methods, apparatus or modules described herein may be implemented in a computer readable program code means for a controller in any suitable manner, for example, the controller may take the form of, for example, a microprocessor or processor and a computer readable medium storing computer readable program code (e.g., software or firmware) executable by the (micro) processor, logic gates, switches, Application Specific Integrated Circuits (ASICs), programmable logic controllers and embedded microcontrollers, examples of which include, but are not limited to, the following microcontrollers: ARC 625D, Atmel AT91SAM, Microchip PIC18F26K20, and Silicone Labs C8051F320, the memory controller may also be implemented as part of the control logic for the memory. Those skilled in the art will also appreciate that, in addition to implementing the controller as pure computer readable program code, the same functionality can be implemented by logically programming method steps such that the controller is in the form of logic gates, switches, application specific integrated circuits, programmable logic controllers, embedded microcontrollers and the like. Such a controller may therefore be considered as a hardware component, and the means included therein for performing the various functions may also be considered as a structure within the hardware component. Or even means for performing the functions may be regarded as being both a software module for performing the method and a structure within a hardware component.
The embodiments in the present specification are described in a progressive manner, and the same or similar parts among the embodiments may be referred to each other, and each embodiment focuses on the differences from the other embodiments. All or portions of the present application are operational with numerous general purpose or special purpose computing system environments or configurations. For example: personal computers, server computers, hand-held or portable devices, tablet-type devices, mobile communication terminals, multiprocessor systems, microprocessor-based systems, programmable electronic devices, network PCs, minicomputers, mainframe computers, distributed computing environments that include any of the above systems or devices, and the like.
The above embodiments are only used to illustrate the technical solutions of the present application, and not to limit the present application; although the present application 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 solutions described in the foregoing embodiments may still be modified, or some or all of the technical features may be equivalently replaced; such modifications or substitutions do not depart from the spirit and scope of the present disclosure.
Claims (10)
1. A small and micro enterprise loan assessment method is characterized by comprising the following steps:
acquiring power utilization running state data and loan data of a small and micro enterprise;
constructing a loan assessment index system of the small and micro enterprise;
weighting each assessment index in the loan assessment index system, and determining an enterprise comprehensive score which is equal to the weighted sum of the assessment indexes;
and setting a monitoring threshold value according to the enterprise comprehensive score so as to monitor the pre-loan risk and the post-loan risk of the small and micro enterprise.
2. The small micro enterprise loan assessment method according to claim 1, wherein the weighting of each assessment index in the loan assessment index system comprises:
and determining the weight of each evaluation index by adopting an analytic hierarchy process or an entropy method.
3. The small micro enterprise loan assessment method according to claim 1, wherein the weighting of each assessment index in the loan assessment index system comprises:
determining a weight of each of the evaluation indicators using a comprehensive weighting method, the function of the comprehensive weighting method being:
wherein, WjWeight value for comprehensive empowerment, W1jDetermination of weight values, W, for analytic hierarchy Process and TOPSIS Process2jThe weight values are determined for the entropy method and the TOPSIS method.
4. The small micro enterprise loan assessment method according to claim 1, wherein the setting of the monitoring threshold according to the enterprise comprehensive score comprises:
and predicting the monthly power consumption of the small and micro enterprises by adopting a gradient lifting tree method, and further setting the monitoring threshold.
5. The small micro enterprise loan assessment method according to any one of claims 1 to 4, further comprising:
associating the electricity utilization running state data with the loan data, and screening fields to form a data width table;
and processing the data on the basis of the data wide table so as to solve the problem of sample imbalance, process abnormal values and fill up missing values.
6. The small micro enterprise loan assessment method according to claim 5, wherein the processing data on the basis of the data wide table comprises:
an undersampling method or a weight-adjusting sampling method is adopted to solve the problem of sample imbalance;
processing abnormal values by adopting a box line method;
and filling up missing values by adopting a linear regression method.
7. The small micro enterprise loan assessment method according to claim 1, wherein the loan assessment index system comprises: payment credit, electricity utilization stability, industry landscape, electricity utilization increase length and default electricity utilization information.
8. A small and micro corporation loan assessment device, comprising:
the data acquisition unit is used for acquiring power utilization running state data and loan data of the small and micro enterprise;
the system construction unit is used for constructing a loan assessment index system of the small and micro enterprise;
the comprehensive scoring unit is used for giving weight to each assessment index in the loan assessment index system and determining an enterprise comprehensive score, and the enterprise comprehensive score is equal to the weighted sum of the assessment indexes;
and the monitoring scheme generating unit is used for setting a monitoring threshold according to the enterprise comprehensive score so as to monitor the pre-loan risk and the post-loan risk of the small and micro enterprise.
9. The small and micro enterprise loan assessment device is characterized by comprising a memory and a processor;
the memory is to store computer-executable instructions;
the processor is configured to execute the computer-executable instructions to implement the method of any of claims 1-7.
10. A computer-readable storage medium having stored thereon executable instructions that, when executed by a computer, are capable of implementing the method of any one of claims 1-7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011278327.7A CN112508679A (en) | 2020-11-16 | 2020-11-16 | Small and micro enterprise loan risk assessment method and device and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011278327.7A CN112508679A (en) | 2020-11-16 | 2020-11-16 | Small and micro enterprise loan risk assessment method and device and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN112508679A true CN112508679A (en) | 2021-03-16 |
Family
ID=74958020
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011278327.7A Pending CN112508679A (en) | 2020-11-16 | 2020-11-16 | Small and micro enterprise loan risk assessment method and device and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112508679A (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113095931A (en) * | 2021-04-25 | 2021-07-09 | 国家电网有限公司 | Post-loan risk monitoring method and device and computer equipment |
| CN113095712A (en) * | 2021-04-25 | 2021-07-09 | 国家电网有限公司 | Enterprise credit granting score obtaining method and device and computer equipment |
| CN113744043A (en) * | 2021-09-03 | 2021-12-03 | 国网江苏省电力有限公司常州供电分公司 | Credit risk assessment method and device for small and micro enterprises and computer equipment |
| CN115034509A (en) * | 2022-06-30 | 2022-09-09 | 中国建设银行股份有限公司 | Data processing method, device, equipment and storage medium |
| CN115941502A (en) * | 2022-11-24 | 2023-04-07 | 中国联合网络通信集团有限公司 | Method and device for determining network operation and maintenance capability |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR100921618B1 (en) * | 2009-04-28 | 2009-10-15 | 기술신용보증기금 | A technology appraisal method for startup company |
| CN106504125A (en) * | 2016-12-27 | 2017-03-15 | 北京中电普华信息技术有限公司 | A kind of method and device for building power customer overall evaluation system |
| CN107194803A (en) * | 2017-05-19 | 2017-09-22 | 南京工业大学 | P2P net loan borrower credit risk assessment device |
| CN109934371A (en) * | 2017-12-18 | 2019-06-25 | 普华讯光（北京）科技有限公司 | The method that solvency risk identification and prediction are carried out to enterprise based on electricity consumption data |
| CN111369171A (en) * | 2020-03-19 | 2020-07-03 | 上海观安信息技术股份有限公司 | User service safety comprehensive risk assessment method based on combined empowerment |
| CN111612323A (en) * | 2020-05-15 | 2020-09-01 | 国网河北省电力有限公司电力科学研究院 | Electric power credit investigation evaluation method based on big data model |
-
2020
- 2020-11-16 CN CN202011278327.7A patent/CN112508679A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR100921618B1 (en) * | 2009-04-28 | 2009-10-15 | 기술신용보증기금 | A technology appraisal method for startup company |
| CN106504125A (en) * | 2016-12-27 | 2017-03-15 | 北京中电普华信息技术有限公司 | A kind of method and device for building power customer overall evaluation system |
| CN107194803A (en) * | 2017-05-19 | 2017-09-22 | 南京工业大学 | P2P net loan borrower credit risk assessment device |
| CN109934371A (en) * | 2017-12-18 | 2019-06-25 | 普华讯光（北京）科技有限公司 | The method that solvency risk identification and prediction are carried out to enterprise based on electricity consumption data |
| CN111369171A (en) * | 2020-03-19 | 2020-07-03 | 上海观安信息技术股份有限公司 | User service safety comprehensive risk assessment method based on combined empowerment |
| CN111612323A (en) * | 2020-05-15 | 2020-09-01 | 国网河北省电力有限公司电力科学研究院 | Electric power credit investigation evaluation method based on big data model |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113095931A (en) * | 2021-04-25 | 2021-07-09 | 国家电网有限公司 | Post-loan risk monitoring method and device and computer equipment |
| CN113095712A (en) * | 2021-04-25 | 2021-07-09 | 国家电网有限公司 | Enterprise credit granting score obtaining method and device and computer equipment |
| CN113744043A (en) * | 2021-09-03 | 2021-12-03 | 国网江苏省电力有限公司常州供电分公司 | Credit risk assessment method and device for small and micro enterprises and computer equipment |
| CN115034509A (en) * | 2022-06-30 | 2022-09-09 | 中国建设银行股份有限公司 | Data processing method, device, equipment and storage medium |
| CN115941502A (en) * | 2022-11-24 | 2023-04-07 | 中国联合网络通信集团有限公司 | Method and device for determining network operation and maintenance capability |
| CN115941502B (en) * | 2022-11-24 | 2024-07-02 | 中国联合网络通信集团有限公司 | Method and device for determining network operation and maintenance capability |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112508679A (en) | Small and micro enterprise loan risk assessment method and device and storage medium | |
| CN107169628B (en) | Power distribution network reliability assessment method based on big data mutual information attribute reduction | |
| CN112632765B (en) | Combat capability assessment method combining weighting method and SEM method | |
| Tanasijević et al. | Study of Dependability Evaluation for Multi‐hierarchical Systems Based on Max–Min Composition | |
| CN115130887B (en) | Reservoir dam environmental impact evaluation method and device, electronic equipment and storage medium | |
| CN114638498A (en) | ESG evaluation method, ESG evaluation system, electronic equipment and storage equipment | |
| CN107590737A (en) | Personal credit scores and credit line measuring method | |
| CN115033591B (en) | Intelligent detection method, system, storage medium and computer equipment for electric charge data abnormality | |
| CN112182056A (en) | Data detection method, device, equipment and storage medium | |
| CN108733864A (en) | A kind of aircraft wing structure Global sensitivity analysis method based on support vector machines | |
| CN113807469A (en) | Multi-energy user value prediction method, device, storage medium and equipment | |
| CN114446019A (en) | Alarm information processing method, device, equipment, storage medium and product | |
| CN114254762A (en) | Interpretable machine learning model construction method and device and computer equipment | |
| CN110796381B (en) | Modeling method and device for wind control model, terminal equipment and medium | |
| Chang | The application of machine learning models in company bankruptcy prediction | |
| CN116149895A (en) | Big data cluster performance prediction method and device and computer equipment | |
| Du et al. | Spatiotemporal dynamics and fitness analysis of global oil market: Based on complex network | |
| CN115204501A (en) | Enterprise evaluation method and device, computer equipment and storage medium | |
| CN115796585A (en) | Enterprise operation risk assessment method and system | |
| CN114092216A (en) | Enterprise credit rating method, apparatus, computer device and storage medium | |
| CN113657501A (en) | Model adaptive training method, apparatus, device, medium, and program product | |
| CN112926803A (en) | Client deposit loss condition prediction method and device based on LSTM network | |
| CN115860273B (en) | Method, apparatus, computer device and storage medium for predicting electric load | |
| CN117557108B (en) | Training method and device for intelligent identification model of power operation risk | |
| Ruan | Customs risk identification and application based on set pair analysis |
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 | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20210316 |