CN112418738A - Staff operation risk prediction method based on logistic regression - Google Patents

Staff operation risk prediction method based on logistic regression Download PDF

Info

Publication number
CN112418738A
CN112418738A CN202011491129.9A CN202011491129A CN112418738A CN 112418738 A CN112418738 A CN 112418738A CN 202011491129 A CN202011491129 A CN 202011491129A CN 112418738 A CN112418738 A CN 112418738A
Authority
CN
China
Prior art keywords
risk
logistic regression
employee
data
operation risk
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.)
Granted
Application number
CN202011491129.9A
Other languages
Chinese (zh)
Other versions
CN112418738B (en
Inventor
向阳
李锦松
颜科琦
陈继春
黄文�
邬小峰
岳雨蒂
程云
黄奕乐
张欣华
崔文军
李威
曾浩
王承林
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Fuhua Rongke Chengdu Technology Co ltd
Bank Of Luzhou Co ltd
Luzhou Laojiao Group Co Ltd
Original Assignee
Fuhua Rongke Chengdu Technology Co ltd
Bank Of Luzhou Co ltd
Luzhou Laojiao Group Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Fuhua Rongke Chengdu Technology Co ltd, Bank Of Luzhou Co ltd, Luzhou Laojiao Group Co Ltd filed Critical Fuhua Rongke Chengdu Technology Co ltd
Priority to CN202011491129.9A priority Critical patent/CN112418738B/en
Publication of CN112418738A publication Critical patent/CN112418738A/en
Application granted granted Critical
Publication of CN112418738B publication Critical patent/CN112418738B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
    • G06Q10/063—Operations research, analysis or management
    • G06Q10/0635—Risk analysis of enterprise or organisation activities
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N20/00—Machine learning
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • 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
    • G06Q40/02—Banking, e.g. interest calculation or account maintenance

Abstract

The invention provides a logistic regression-based employee operation risk prediction method, which comprises the following steps: s1, collecting operation risk point list data and employee information data; s2, classifying the risk points in the operation risk point list data; s3, extracting operation risk events and employee information data of various risk points to form an operation risk data set; s4, labeling the staff according to the occurrence condition of the operation risk event; s5, respectively carrying out data preprocessing on employee information data in each type of risk point operation risk data set; s6, respectively establishing a logistic regression model for each type of risk points based on the employee labels; s7, training a corresponding logistic regression model by using employee information data in each type of risk point operation risk data set after data preprocessing; and S8, calculating the probability of the employee to have various operation risk events by using the trained logistic regression model. The invention collects the information data of the staff, monitors the behavior of the staff, does not need manual on-site inspection, and reduces the labor cost.

Description

Staff operation risk prediction method based on logistic regression
Technical Field
The invention relates to the field of prevention and control management of operational risks of financial institutions and banking industries, in particular to a staff operational risk prediction method based on logistic regression.
Background
At present, the multiple troubleshooting of financial institutions and banking industries on employee operation risk behaviors is limited to manual troubleshooting and supervision, and a large amount of manpower and material resources are consumed. The scope of investigation and supervision is often limited to only individual departments or branches and cannot cover the entire institution or bank. The internal control mechanism and the supervision mechanism of the bank have disadvantages, the management and control of the operation risk are thoroughly checked after the incident, the advance prevention is carried out, a plurality of events are investigated in detail only after the risk occurs, the advance prevention work is omitted, and the mechanism misses the risk in the intangible opportunity or the opportunity of reducing the risk cost, so that unnecessary loss is caused.
Disclosure of Invention
The invention aims to provide a method for predicting the operation risk of staff based on logistic regression, which aims to solve the problems of manual investigation and supervision of the operation risk of the staff.
The invention provides a logistic regression-based employee operation risk prediction method, which comprises the following steps:
s1, collecting operation risk point list data and employee information data;
s2, classifying the risk points in the operation risk point list data according to the influence degree;
s3, extracting operation risk events and employee information data of various risk points to form an operation risk data set;
s4, labeling the staff according to the occurrence condition of the operation risk event in each type of risk point operation risk data set;
s5, respectively carrying out data preprocessing on employee information data in each type of risk point operation risk data set;
s6, respectively establishing a logistic regression model for each type of risk points based on the employee labels;
s7, training a corresponding logistic regression model by using employee information data in each type of risk point operation risk data set after data preprocessing;
and S8, calculating the probability of the employee to have various operation risk events by using the trained logistic regression model.
Further, the method for labeling the employee in step S4 includes: the label of the operation risk event of the employee who has occurred the operation risk event of the risk point at the risk point is 1, and the label of the operation risk event of the employee who has not occurred the operation risk event of the risk point at the risk point is 0.
Further, the data preprocessing in step S5 includes:
data cleaning: abnormal value processing, missing value processing, repeated value deleting, qualitative variable conversion into numerical variable, and date conversion from text format to time format;
generating a derivative variable: and (4) taking the employee information data subjected to data cleaning as an original variable, and designing a derivative condition to obtain a derivative variable of the original variable.
Further, the method for establishing the logistic regression model in step S6 is to map the result to the space between (0,1) through a Sigmoid function on the basis of linear regression, thereby obtaining a logistic regression model for calculating the probability of the employee to have a certain operation risk event; the expression of the logistic regression model is:
Figure BDA0002840714070000021
wherein:
Figure BDA0002840714070000022
representing the probability of the nth worker with the label of 1 to have certain operation risk event;
Figure BDA0002840714070000023
Figure BDA0002840714070000024
the method comprises the steps of inputting data of a logistic regression model, namely original variables and derivative variables of employee information data of a risk point category; w is aTThe optimal coefficient obtained by the loss function training is obtained; x is the number ofnFirst to represent the category of the risk points to which it belongsOriginal variables and derivative variables of employee information data of n employees; w is anDenotes xnThe optimum coefficient of (a); y isnA label representing the nth employee of the category of risk points.
Further, the loss function is:
Figure BDA0002840714070000025
where N represents the total number of employees in the risk point category to which they belong.
Further, step S7 includes the following sub-steps:
s71, dividing original variables and derived variables of employee information data of the risk point category into a training set and a test set;
s72, performing variable primary screening on the training set by adopting correlation analysis and variance expansion factors;
s73, calculating woe values and iv values, and selecting input data to be input into the logistic regression model from the initially screened training set according to the calculated iv values;
s74, inputting the selected input data into the logistic regression model, and training the logistic regression model by using a loss function;
s75, evaluating the logistic regression model after the test set input training, wherein the evaluated logistic regression model is the trained logistic regression model;
and S76, when the collected operation risk point list data and the staff information data are updated, the logistic regression model is updated accordingly.
Further, step S8 includes setting a probability threshold, and performing early warning by comparing the probability threshold with the calculated probability of various operation risk events of the employee.
Further, the probability threshold is multiple, and each probability threshold represents a different risk influence degree.
Further, the early warning is pushed to a manager through a platform message or a short message.
Further, the prediction method further includes:
and S9, collecting and counting the actual situations of various operation risk events of the staff, comparing the calculated probability of various operation risk events of the staff with the actual situations, and optimizing the logistic regression model according to the comparison result.
In summary, due to the adoption of the technical scheme, the invention has the beneficial effects that:
1. the invention collects the information data of the staff, monitors the behavior of the staff, does not need manual on-site inspection, and reduces the labor cost.
2. According to the invention, a machine learning algorithm is introduced, and a model is established through objective data, so that early warning can be given before an operation risk event occurs, high-risk staff are focused, the energy dispersion is avoided, and unnecessary loss is reduced.
3. The invention automatically updates the model according to the data, monitors the model performance in real time and is convenient for manual intervention to modify the model in time.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings in the embodiments will be briefly described below, it should be understood that the following drawings only illustrate some embodiments of the present invention, and therefore should not be considered as limiting the scope, and for those skilled in the art, other related drawings can be obtained according to the drawings without inventive efforts.
Fig. 1 is a flowchart of an employee operational risk prediction method based on logistic regression according to an embodiment of the present invention.
FIG. 2 is a block diagram of a process for training a logistic regression model according to an embodiment of the present invention.
FIG. 3 is a block flow diagram of a method for predicting employee operational risk based on logistic regression according to another embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are some, but not all, embodiments of the present invention. The components of embodiments of the present invention generally described and illustrated in the figures herein may be arranged and designed in a wide variety of different configurations.
Thus, the following detailed description of the embodiments of the present invention, presented in the figures, is not intended to limit the scope of the invention, as claimed, but is merely representative of selected embodiments of the invention. 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.
Examples
As shown in fig. 1, the embodiment provides a method for predicting employee operational risk based on logistic regression, which includes the following steps:
s1, collecting operation risk point list data and employee information data;
the employee information data comprises personal information data and working behavior data, and the main system sources for collecting the data comprise an operation risk system, an OA system and various business systems, such as: the financial management system, the counter system, the credit system and the like take the acquired data as the basis of subsequent analysis, and are favorable for improving the prediction precision. In this embodiment, the employee information data includes the following collection ranges:
a. basic information, including: gender, age, education level, department of the department, post, number of leave requests, leave requests duration, length of rest, etc.;
b. work information, including: work content and work behavior: login, operation condition, task number, completion condition and task completion time consumption in each system; responsibility authority, work capacity; work communication, work diary: the working circle information updating time, the number of the working circle information pieces, the working circle information content, the communication condition, the working log updating times, the working log updating content, the announcement reading time length, the announcement reading times and the like;
c. fund traffic information, including: employee deposit accounts, transaction contents, transaction opponents, transaction amount, transaction time, employee investment account numbers, investment amount, investment time and the like;
d. operational risk information, including: operational risk events occurring by employees in the operational risk event repository, risk event severity, and the like.
The above listed metrics may be divided according to differences in content, whether quantitative or qualitative. The quantitative data is mainly obtained by analyzing historical data, and analysis dimensions such as work completion degree, work total amount and the like are displayed in a numerical value mode.
S2, classifying the risk points in the operation risk point list data according to the influence degree;
that is, the operation risk point list in the mechanism is combed by the service experts in the mechanism or in the line, and the risk points are classified according to the influence degree of each risk point, wherein the influence degree comprises: relating to monetary amounts, legal impacts, reputation impacts, relating to scope, risk content, and the like.
S3, extracting operation risk events and employee information data of various risk points to form an operation risk data set;
that is, the operation risk data sets of the respective classifications are formed according to the risk point classifications, and the operation risk events of the risk points and the employee information data related to the risk points are included in the operation risk data sets.
S4, labeling the staff according to the occurrence condition of the operation risk event in each type of risk point operation risk data set;
specifically, the method for labeling the staff comprises the following steps: the label of the operation risk event of the employee who has occurred the operation risk event of the risk point at the risk point is 1, and the label of the operation risk event of the employee who has not occurred the operation risk event of the risk point at the risk point is 0.
S5, respectively carrying out data preprocessing on employee information data in each type of risk point operation risk data set;
the data preprocessing comprises the following steps:
data cleaning: outlier processing (e.g., employee age 100 delete), missing value processing, duplicate value delete, qualitative variable conversion to numerical variable (e.g., academic high school and 1 below, major 2, subject 3, master 4, doctor 5), and date conversion from text format to time format. Generally, the code for data cleaning is abstracted and packaged into a function, so that the function is convenient to call in a system.
Generating a derivative variable: and (4) taking the employee information data subjected to data cleaning as an original variable, and designing a derivative condition to obtain a derivative variable of the original variable. The derived variables can be created by business personnel from business experience in the system derived variable creation page. The derived variable creation page lists all original variables, and performs a combination operation on the existing original variables, such as: plus, -x,/, etc. to generate new variables, i.e. derived variables. The derived variable calculation formula can be generated according to the derived conditions by designing the derived conditions, such as: the job post is a teller, the workload formula is the number of accounts processed, the job post is a credit employee, and the workload is the loan amount.
S6, respectively establishing a logistic regression model for each type of risk points based on the employee labels;
the method for establishing the logistic regression model is that the result is mapped between (0,1) through a Sigmoid function on the basis of linear regression, so that the logistic regression model for calculating the probability of the occurrence of certain operation risk events of the staff is obtained;
the Sigmoid function is a step function. By ynThe n-th employee's label representing the category of the risk point is that the label of the operation risk event of the employee who has occurred the operation risk event of the risk point at the risk point is 1, and the label of the operation risk event of the employee who has not occurred the operation risk event of the risk point at the risk point is 0, that is, ynE (0, 1). Regarding the probability of the nth worker occurring the operation risk event as pnThen the label is 1 (i.e., y)nThe probability of an operation risk event occurring for an employee of 1) is:
Figure BDA0002840714070000061
wherein the content of the first and second substances,
Figure BDA0002840714070000062
Figure BDA0002840714070000063
the input data of the logistic regression model, namely the original variables and the derivative variables of the employee information data of the risk point category, for example: the operation risks of large amount of capital and currency between the staff and the client are caused, and the data of the operation risks comprise the currency amount, the currency frequency and the like; w is aTThe optimal coefficient obtained by the loss function training is obtained; x is the number ofnThe original variable and the derivative variable of the employee information data of the nth employee of the risk point category are represented; w is anDenotes xnThe optimum coefficient of (c). When z approaches positive infinity, the probability pnApproaching 1, when z approaches negative infinity, pnApproaching 0. The independent variable is taken as any real number, a predicted value z can be obtained through linear regression, and the value is mapped to a Sigmoid function, so that the conversion from the value to the probability is completed.
Similarly, the label is 0 (i.e., y)n0) is given by:
Figure BDA0002840714070000071
thus, the expression of the logistic regression model can be written as:
Figure BDA0002840714070000072
for the loss function, assuming that the employees are independent of each other, the probability of the operational risk event occurring according to the maximum likelihood function set of input data can be written as:
Figure BDA0002840714070000073
the logarithm of the likelihood function is taken, the monotonicity of the original function is not influenced by the logarithm, the difference between the probabilities can be amplified, and the category of each sample can be better distinguished. To convert to:
Figure BDA0002840714070000074
then at log PGeneral assemblyThe addition of a negative sign in front becomes the minimum negative log-likelihood function:
Figure BDA0002840714070000075
minimizing the negative log-likelihood function averages N samples (N also represents the total number of employees in the risk point category) to obtain a loss function:
Figure BDA0002840714070000076
the loss function is thus mapped to the distribution law of the 0-1 distribution by taking the logarithm on the basis of the 0-1 distribution and then taking the negative number. When y isnWhen 1, pnThe closer to 1, the smaller the loss function; when y isnWhen equal to 0, pnThe closer to 0, the smaller the loss function. Thus, by training, p can be forcednApproach to ynAnd thus correctly classified.
S7, training a corresponding logistic regression model by using employee information data in each type of risk point operation risk data set after data preprocessing;
specifically, step S7 includes the following sub-steps:
s71, dividing original variables and derived variables of employee information data of the risk point category into a training set and a test set;
s72, performing variable primary screening on the training set by adopting correlation analysis and variance expansion factors;
s73, calculating woe values and iv values, and selecting input data to be input into the logistic regression model from the initially screened training set according to the calculated iv values;
the continuous variables are firstly subjected to binning (namely segmentation) discretization by using an automatic binning function, boxes with similar continuous variables and class boxes with similar continuous variables are merged and distributed in the automatic binning function by using chi-square test, and the number of the boxes, the box subintervals and the woe values of all the variables are determined. The system gives the number of recommended boxes, and the customer can adjust the number of the boxes of the variable according to the staff distribution histogram of the variable given by the system interface. The iv values for all variables are then calculated from the calculated woe values, and the iv values select the input data to be input into the logistic regression model from the initially screened training set. Wherein, an iv threshold value can be set to screen out variables with large iv value (such as iv > 0.03) and recommend the variables to the user, and then the user selects the variables with large iv value from the screened variables as the input data of the logistic regression model.
S74, inputting the selected input data into the logistic regression model, and training the logistic regression model by using a loss function; i.e. w can be obtained using a loss functionnTo thereby complete model training.
S75, evaluating the logistic regression model after the test set input training, wherein the evaluated logistic regression model is the trained logistic regression model; the test set may also be screened in steps S72 and S73 and then input into a trained logistic regression model for evaluation. And displaying the evaluation result to a user through a graphical interface, and if the user judges that the evaluation result is not satisfactory, reselecting the variable to train the logistic regression model.
And S76, when the collected operation risk point list data and the staff information data are updated, the logistic regression model is updated accordingly. That is, as the collected data is accumulated, the logistic regression model should be updated to be closer to the current decision result. On the other hand, the logistic regression model may be set to be updated periodically in view of saving computing resources.
And S8, calculating the probability of the employee to have various operation risk events by using the trained logistic regression model. And each type of risk point has a trained logistic regression model, and the probability of various operation risk events of the staff can be predicted by applying the logistic regression models of various types of risk points. Further, step S8 includes setting a probability threshold, and performing early warning by comparing the probability threshold with the calculated probability of various operation risk events of the employee. The early warning is pushed to a manager through platform messages or short messages, and the manager is informed to pay key attention in time. Still further, there are a plurality of probability threshold values, represent different risk influence degree respectively, can let managers know the risk level directly perceivedly from this.
In some embodiments, the prediction method further comprises:
and S9, collecting and counting the actual situations of various operation risk events of the staff, comparing the calculated probability of various operation risk events of the staff with the actual situations, and optimizing the logistic regression model according to the comparison result.
Monitoring the prediction effect of the current logistic regression model, wherein the indexes comprise: the model distinguishing capacity ks value, the model prediction accuracy, the model stability PSI, the model misplacement rate (the proportion of the employees not exceeding the early warning probability threshold value that the operation risk events actually occur), the good employee prediction accuracy rate (the proportion of the employees not exceeding the early warning probability threshold value that the operation risk events actually occur) and the bad employee prediction accuracy rate (the proportion of the employees not exceeding the early warning probability threshold value that the operation risk events actually occur). The calculation formulas of the good employee prediction accuracy and the bad employee prediction accuracy are obtained based on the actual conditions of the project. And continuously monitoring the logistic regression model, comparing the change trends of the model at different periods through a line graph, and timely adjusting the model when the effect of the model is reduced.
The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention, and various modifications and changes may be made by those skilled in the art. Any modification, equivalent replacement, or improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims (10)

1. A staff operation risk prediction method based on logistic regression is characterized by comprising the following steps:
s1, collecting operation risk point list data and employee information data;
s2, classifying the risk points in the operation risk point list data according to the influence degree;
s3, extracting operation risk events and employee information data of various risk points to form an operation risk data set;
s4, labeling the staff according to the occurrence condition of the operation risk event in each type of risk point operation risk data set;
s5, respectively carrying out data preprocessing on employee information data in each type of risk point operation risk data set;
s6, respectively establishing a logistic regression model for each type of risk points based on the employee labels;
s7, training a corresponding logistic regression model by using employee information data in each type of risk point operation risk data set after data preprocessing;
and S8, calculating the probability of the employee to have various operation risk events by using the trained logistic regression model.
2. The method for predicting the operational risk of the employee based on the logistic regression as claimed in claim 1, wherein the method for labeling the employee in the step S4 is as follows: the label of the operation risk event of the employee who has occurred the operation risk event of the risk point at the risk point is 1, and the label of the operation risk event of the employee who has not occurred the operation risk event of the risk point at the risk point is 0.
3. The method for predicting the operational risk of the employee based on the logistic regression as claimed in claim 1, wherein the data preprocessing in the step S5 comprises:
data cleaning: abnormal value processing, missing value processing, repeated value deleting, qualitative variable conversion into numerical variable, and date conversion from text format to time format;
generating a derivative variable: and (4) taking the employee information data subjected to data cleaning as an original variable, and designing a derivative condition to obtain a derivative variable of the original variable.
4. The method for predicting the operational risk of the employee based on the logistic regression as claimed in claim 3, wherein the logistic regression model is established in step S6 by mapping the result between (0,1) through a Sigmoid function on the basis of the linear regression so as to obtain the logistic regression model for calculating the probability of the employee to have a certain operational risk event; the expression of the logistic regression model is:
Figure FDA0002840714060000021
wherein:
Figure FDA0002840714060000022
representing the probability of the nth worker with the label of 1 to have certain operation risk event;
Figure FDA0002840714060000023
Figure FDA0002840714060000024
the method comprises the steps of inputting data of a logistic regression model, namely original variables and derivative variables of employee information data of a risk point category; w is aTThe optimal coefficient obtained by the loss function training is obtained; x is the number ofnThe original variable and the derivative variable of the employee information data of the nth employee of the risk point category are represented; w is anDenotes xnThe optimum coefficient of (a); y isnA label representing the nth employee of the category of risk points.
5. The method for predicting the operational risk of an employee based on logistic regression as recited in claim 4, wherein said loss function is:
Figure FDA0002840714060000025
where N represents the total number of employees in the risk point category to which they belong.
6. The method for predicting the operational risk of the employee based on the logistic regression as claimed in claim 4 or 5, wherein the step S7 comprises the following sub-steps:
s71, dividing original variables and derived variables of employee information data of the risk point category into a training set and a test set;
s72, performing variable primary screening on the training set by adopting correlation analysis and variance expansion factors;
s73, calculating woe values and iv values, and selecting input data to be input into the logistic regression model from the initially screened training set according to the calculated iv values;
s74, inputting the selected input data into the logistic regression model, and training the logistic regression model by using a loss function;
s75, evaluating the logistic regression model after the test set input training, wherein the evaluated logistic regression model is the trained logistic regression model;
and S76, when the collected operation risk point list data and the staff information data are updated, the logistic regression model is updated accordingly.
7. The method for predicting the operational risk of the employee based on the logistic regression as claimed in claim 1, wherein the step S8 further comprises setting a probability threshold, and performing the early warning by comparing the probability threshold with the calculated probability of various operational risk events of the employee.
8. The method for predicting the operational risk of the employee based on the logistic regression as claimed in claim 7, wherein the probability threshold is plural and represents different degrees of risk influence.
9. The method for predicting the operational risk of the employee based on the logistic regression as claimed in claim 7 or 8, wherein the early warning is pushed to the manager through a platform message or a short message.
10. The method for forecasting employee operational risk based on logistic regression as claimed in claim 7 or 8, characterized in that the forecasting method further comprises:
and S9, collecting and counting the actual situations of various operation risk events of the staff, comparing the calculated probability of various operation risk events of the staff with the actual situations, and optimizing the logistic regression model according to the comparison result.
CN202011491129.9A 2020-12-17 2020-12-17 Staff operation risk prediction method based on logistic regression Active CN112418738B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202011491129.9A CN112418738B (en) 2020-12-17 2020-12-17 Staff operation risk prediction method based on logistic regression

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202011491129.9A CN112418738B (en) 2020-12-17 2020-12-17 Staff operation risk prediction method based on logistic regression

Publications (2)

Publication Number Publication Date
CN112418738A true CN112418738A (en) 2021-02-26
CN112418738B CN112418738B (en) 2022-02-11

Family

ID=74775686

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202011491129.9A Active CN112418738B (en) 2020-12-17 2020-12-17 Staff operation risk prediction method based on logistic regression

Country Status (1)

Country Link
CN (1) CN112418738B (en)

Cited By (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113128760A (en) * 2021-04-16 2021-07-16 深圳市艾赛克科技有限公司 Double-control management method, device, equipment and storage medium for safety production
CN115408379A (en) * 2022-10-25 2022-11-29 广州市玄武无线科技股份有限公司 Terminal repeating data determination method, device, equipment and computer storage medium
CN116070725A (en) * 2022-08-29 2023-05-05 山东科技大学 Mining pressure risk prediction method based on logistic regression
CN117251851A (en) * 2023-11-03 2023-12-19 广东齐思达信息科技有限公司 Internet surfing behavior management auditing method
CN117251851B (en) * 2023-11-03 2024-05-14 广东齐思达信息科技有限公司 Internet surfing behavior management auditing method

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20140297361A1 (en) * 2012-07-12 2014-10-02 Bank Of America Corporation Operational risk back-testing process using quantitative methods
CN106157132A (en) * 2016-06-20 2016-11-23 中国工商银行股份有限公司 Credit risk monitoring system and method
CN108492049A (en) * 2018-04-04 2018-09-04 冯世程 A kind of system for the P2P platform operation risk assessment that logic-based returns
CN109299740A (en) * 2018-09-28 2019-02-01 北京赛博贝斯数据科技有限责任公司 A kind of construction method and system of Logic Regression Models
CN109508825A (en) * 2018-11-12 2019-03-22 平安科技（深圳）有限公司 Employee's image method for prewarning risk and relevant apparatus
CN109615280A (en) * 2019-01-25 2019-04-12 平安科技（深圳）有限公司 Employee's data processing method, device, computer equipment and storage medium

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20140297361A1 (en) * 2012-07-12 2014-10-02 Bank Of America Corporation Operational risk back-testing process using quantitative methods
CN106157132A (en) * 2016-06-20 2016-11-23 中国工商银行股份有限公司 Credit risk monitoring system and method
CN108492049A (en) * 2018-04-04 2018-09-04 冯世程 A kind of system for the P2P platform operation risk assessment that logic-based returns
CN109299740A (en) * 2018-09-28 2019-02-01 北京赛博贝斯数据科技有限责任公司 A kind of construction method and system of Logic Regression Models
CN109508825A (en) * 2018-11-12 2019-03-22 平安科技（深圳）有限公司 Employee's image method for prewarning risk and relevant apparatus
CN109615280A (en) * 2019-01-25 2019-04-12 平安科技（深圳）有限公司 Employee's data processing method, device, computer equipment and storage medium

Cited By (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113128760A (en) * 2021-04-16 2021-07-16 深圳市艾赛克科技有限公司 Double-control management method, device, equipment and storage medium for safety production
CN116070725A (en) * 2022-08-29 2023-05-05 山东科技大学 Mining pressure risk prediction method based on logistic regression
CN115408379A (en) * 2022-10-25 2022-11-29 广州市玄武无线科技股份有限公司 Terminal repeating data determination method, device, equipment and computer storage medium
CN117251851A (en) * 2023-11-03 2023-12-19 广东齐思达信息科技有限公司 Internet surfing behavior management auditing method
CN117251851B (en) * 2023-11-03 2024-05-14 广东齐思达信息科技有限公司 Internet surfing behavior management auditing method

Also Published As

Publication number Publication date
CN112418738B (en) 2022-02-11

Similar Documents

Publication Publication Date Title
CN112418738B (en) Staff operation risk prediction method based on logistic regression
Ganesh et al. Future of artificial intelligence and its influence on supply chain risk management–A systematic review
US11037080B2 (en) Operational process anomaly detection
US8214308B2 (en) Computer-implemented systems and methods for updating predictive models
CN109191282A (en) Methods of marking and system are monitored in a kind of loan of Behavior-based control model
Kuenkaikaew Predictive Audit Analytics: Evolving to a new era
CN112700319A (en) Enterprise credit line determination method and device based on government affair data
CN114048436A (en) Construction method and construction device for forecasting enterprise financial data model
CN113537807B (en) Intelligent wind control method and equipment for enterprises
Dbouk et al. Towards a machine learning approach for earnings manipulation detection
Wang et al. Integrating decision tree with back propagation network to conduct business diagnosis and performance simulation for solar companies
CN114078050A (en) Loan overdue prediction method and device, electronic equipment and computer readable medium
CN117151345A (en) Enterprise management intelligent decision platform based on AI technology
CN110689437A (en) Communication construction project financial risk prediction method based on random forest
CN114154866A (en) Marketing enterprise financial risk early warning method and system
CN112686446A (en) Machine learning interpretability-oriented credit default prediction method and system
Scherger et al. DETECTION AND ASSESSMENT OF CAUSES IN BUSINESS DIAGNOSIS.
CN116228403A (en) Personal bad asset valuation method and system based on machine learning algorithm
Koyuncugil et al. Early warning system for financially distressed hospitals via data mining application
KR102499182B1 (en) Loan regular auditing system using artificia intellicence
CN115689713A (en) Abnormal risk data processing method and device, computer equipment and storage medium
KR102499181B1 (en) Loan regular auditing system using artificia intellicence
CN114612239A (en) Stock public opinion monitoring and wind control system based on algorithm, big data and artificial intelligence
CN114219606A (en) Power data-based power consumption enterprise credit evaluation method and system
Jadav et al. EmReSys: AI-based Efficient Employee Ranking and Recommender System for Organizations

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant