CN114997512A

CN114997512A - Employee departure prediction method, device, equipment and storage medium

Info

Publication number: CN114997512A
Application number: CN202210728027.7A
Authority: CN
Inventors: 黎伟浚; 巴堃; 庄伯金
Original assignee: Ping An Technology Shenzhen Co Ltd
Current assignee: Ping An Technology Shenzhen Co Ltd
Priority date: 2022-06-23
Filing date: 2022-06-23
Publication date: 2022-09-02

Abstract

The invention relates to an artificial intelligence technology, and discloses an employee departure prediction method, which comprises the following steps: acquiring a characteristic data set under a working scene; filling missing values in the feature data set to obtain a standard feature data set, wherein the standard feature data set comprises a static feature data set and a dynamic feature data set; an initial prediction model is built by using the static characteristic data set, and the initial probability of leaving the job of the characteristic data set is calculated by using the initial prediction model to obtain an initial probability set; extracting time sequence characteristics in the dynamic characteristic data set to obtain a time sequence characteristic set; and constructing an employee job leaving prediction model by utilizing the time sequence characteristic set and the initial probability set, and calculating the employee job leaving probability by utilizing the job leaving prediction model. The invention also provides an employee attendance prediction system device, electronic equipment and a storage medium. The invention can improve the accuracy of employee departure prediction.

Description

Employee departure prediction method, device, equipment and storage medium

Technical Field

The invention relates to the technical field of artificial intelligence, in particular to a staff position-leaving prediction method and device, electronic equipment and a computer-readable storage medium.

Background

For enterprises, intervention is performed after employees propose a leave application behavior, so that talents are difficult to effectively save excellent employees and cause talents loss; and the behavior of the staff with high job leaving risk on the confidential information is uncontrollable, and the risk of illegal disclosure of the confidential information is easy to exist, so that the enterprise is very necessary to perfect the prediction of the job leaving risk of the staff.

Most of the characteristics considered by the traditional employee job leaving prediction scheme are static characteristics of employees, for example, basic information, historical work information, job level and performance information of the employees and the like, only a small amount of dynamic characteristics exist, but the employees with higher risk level usually have changes of recent behavior dynamic characteristics, meanwhile, for different business posts, business characteristics with strong correlation with job leaving risks are unique, the existing technical scheme does not subdivide the characteristics of the employees for modeling, so that the characteristic data of employee job leaving risk prediction is not comprehensive, and the accuracy of employee job leaving risk prediction is low.

Disclosure of Invention

The invention provides a staff leave prediction method, a device and a computer readable storage medium, and mainly aims to solve the problem of low staff leave prediction accuracy.

In order to achieve the above object, the present invention provides a method for predicting employee leave, comprising:

acquiring feature labels of enterprise employees in different fields in an enterprise under a working scene, and acquiring a feature data set under the working scene according to the feature labels;

filling missing values into the feature data set to obtain a standard feature data set, wherein the feature data set comprises a static feature data set and a dynamic feature data set;

constructing an initial prediction model by using the static characteristic data set, and calculating an initial job leaving probability of the characteristic data set by using the initial prediction model to obtain an initial probability set;

extracting time sequence characteristics in the dynamic characteristic data set by using a preset time sequence characteristic model to obtain a time sequence characteristic set;

and constructing an employee job leaving prediction model according to the time sequence characteristic set and the initial probability set, and calculating the employee job leaving probability by using the job leaving prediction model.

Optionally, the obtaining a feature data set in the working scenario according to the feature tag includes:

acquiring feature attribute information corresponding to the feature tags, and determining extraction tags of feature data according to the feature attribute information;

and extracting feature data from a pre-constructed employee database based on the extraction label to obtain a feature data set under the working scene.

Optionally, the missing value filling the feature data set to obtain a standard feature data set includes:

searching missing value data labels of each employee data in the characteristic data set according to the characteristic labels of the characteristic data set;

extracting a missing data set corresponding to the missing value data label from the characteristic data set;

and calculating a missing value of each employee data based on the missing data set, and filling the missing values into the feature data set to obtain a standard feature data set.

Optionally, the constructing an initial prediction model by using the static feature data set includes:

carrying out normalization processing on the static characteristic data set;

performing model training on a pre-constructed XGboost algorithm model based on the static feature data set after normalization processing to obtain an output probability value;

calculating an optimal objective function of the XGboost algorithm model according to the output probability value, and determining an optimal decision tree in the XGboost algorithm model according to the optimal objective function;

and determining the optimal parameters of the XGboost algorithm model according to the optimal decision tree to obtain an initial prediction model.

Optionally, the calculating an optimal objective function of the XGBoost algorithm model according to the output probability value includes:

calculating a minimum loss function and a regularization function of the XGboost algorithm model according to the output probability value, and adding the minimum loss function and the regularization function to obtain a target function;

performing second-order Taylor expansion on the target function to obtain a Taylor expansion, and removing a constant term in the Taylor expansion to obtain a function expansion;

and combining the first term coefficient and the second term coefficient in the function expansion to obtain an optimal objective function.

Optionally, the extracting, by using a preset time series feature model, the time series feature in the dynamic feature data set to obtain a time series feature set includes:

performing convolution on the dynamic characteristic data by utilizing a convolution layer in the time sequence characteristic model to obtain a convolution characteristic diagram, and performing space pooling on the convolution characteristic diagram;

performing convolution channel compression on the convolution characteristic graph after the space pooling by using a convolution core in the convolution layer to obtain a compressed characteristic graph;

and performing expansion convolution operation on the compression characteristic diagram through the convolution layer to obtain the time sequence characteristics of the characteristic data set.

Optionally, the performing, by the convolutional layer, a dilation convolution operation on the compressed feature map includes:

performing a dilation convolution operation on the compressed feature map by using the following formula:

wherein K is the number of convolution layers, f _k For the kth convolutional layer, X is the sequence of the expansion convolution, X _t For the t-th dilation convolution operation, d is a dilation parameter,

is the x _t And (4) calculating the time sequence characteristics by the dilation convolution.

In order to solve the above problems, the present invention also provides an employee departure prediction apparatus, including:

the system comprises a characteristic data acquisition module, a characteristic data acquisition module and a characteristic data acquisition module, wherein the characteristic data acquisition module is used for acquiring characteristic labels of enterprise employees in different fields in an enterprise under a working scene and acquiring a characteristic data set under the working scene according to the characteristic labels;

a missing value filling module, configured to perform missing value filling on the feature data set to obtain a standard feature data set, where the feature data set includes a static feature data set and a dynamic feature data set;

the initial prediction module is used for constructing an initial prediction model by using the static characteristic data set, and calculating the initial job leaving probability of the characteristic data set by using the initial prediction model to obtain an initial probability set;

the time sequence feature extraction module is used for extracting time sequence features in the dynamic feature data set by using a preset time sequence feature model to obtain a time sequence feature set;

and the employee departure prediction model building module is used for building an employee departure prediction model according to the time sequence characteristic set and the initial probability set and calculating the employee departure probability by using the departure prediction model.

In order to solve the above problem, the present invention also provides an electronic device, including:

at least one processor; and (c) a second step of,

a memory communicatively coupled to the at least one processor; wherein the content of the first and second substances,

the memory stores a computer program executable by the at least one processor, the computer program being executable by the at least one processor to enable the at least one processor to perform the employee departure prediction method described above.

In order to solve the above problem, the present invention further provides a computer-readable storage medium, in which at least one computer program is stored, and the at least one computer program is executed by a processor in an electronic device to implement the employee attendance prediction method described above.

The embodiment of the invention further acquires the characteristic data by acquiring the characteristic label in the working scene, so that the characteristic data comprises the characteristic data specific to the working scene, the characteristic data is more comprehensive, and the prediction accuracy of the risk behavior prediction model is favorably improved; and filling missing values in the feature data to obtain a standard feature data set, wherein the feature data set comprises a static feature data set and a dynamic feature data set. Subdividing the characteristic data set and modeling step by step to improve the prediction accuracy of the employee job leaving prediction model; constructing an initial prediction model by using the static characteristic data set to carry out risk prediction on the characteristic data set to obtain an initial probability set; extracting time sequence characteristics in the dynamic characteristic data set, and further enlarging the range of the characteristics; and constructing an employee job leaving prediction model by utilizing the time sequence characteristic set and the initial probability set, and calculating the employee job leaving probability by utilizing the job leaving prediction model so as to improve the prediction accuracy of employee job leaving prediction. Therefore, the employee departure prediction method, the employee departure prediction device, the electronic equipment and the computer readable storage medium can solve the problem of low accuracy in employee departure prediction.

Drawings

Fig. 1 is a schematic flowchart of an employee job departure prediction method according to an embodiment of the present invention;

FIG. 2 is a schematic flow chart of constructing an initial prediction model according to an embodiment of the present invention;

fig. 3 is a schematic flow chart illustrating a process of extracting timing characteristics according to an embodiment of the present invention;

fig. 4 is a functional block diagram of an employee departure prediction apparatus according to an embodiment of the present invention;

fig. 5 is a schematic structural diagram of an electronic device for implementing the employee departure prediction method according to an embodiment of the present invention.

The implementation, functional features and advantages of the objects of the present invention will be further explained with reference to the accompanying drawings.

Detailed Description

It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

The embodiment of the application provides a staff leave prediction method. The execution subject of the employee attendance prediction method includes, but is not limited to, at least one of electronic devices such as a server and a terminal, which can be configured to execute the method provided by the embodiment of the present application. In other words, the employee attendance prediction method may be executed by software or hardware installed in the terminal device or the server device, and the software may be a block chain platform. The server includes but is not limited to: a single server, a server cluster, a cloud server or a cloud server cluster, and the like. The server may be an independent server, or may be a cloud server that provides basic cloud computing services such as a cloud service, a cloud database, cloud computing, a cloud function, cloud storage, a Network service, cloud communication, a middleware service, a domain name service, a security service, a Content Delivery Network (CDN), a big data and artificial intelligence platform, and the like.

Fig. 1 is a schematic flow chart of an employee attendance prediction method according to an embodiment of the present invention. In this embodiment, the employee departure prediction method includes:

s1, acquiring feature labels of enterprise employees in different fields in an enterprise in a working scene, and acquiring a feature data set in the working scene according to the feature labels;

in the embodiment of the present invention, the work scene may include, for example, a work scene of a development post of an enterprise, a work scene of a business post, and the like, and different business feature tags are provided in different work scenes, for example, the development post has a code submission feature tag, and the business post has an outwork feature tag, and the like.

In detail, the obtaining of the feature data set under the working scenario according to the feature tag includes:

In the embodiment of the present invention, the feature attribute information may include a plurality of feature information, for example, employee age information, employee title information, and employee work change information of the employee, and an extraction label of data is obtained based on the feature information, for example, rules such as not extracting employee feature data whose employee age is less than six months are used.

In the embodiment of the invention, the characteristic data set has the service characteristic data under different working scenes by acquiring the characteristic data set under the working scenes, so that the characteristic data is more comprehensive, and meanwhile, the method has stronger pertinence to the staff under different working scenes, and the prediction of the staff behavior risk is more accurate.

S2, filling missing values in the feature data set to obtain a standard feature data set, wherein the feature data set comprises a static feature data set and a dynamic feature data set;

in the embodiment of the invention, because the characteristic data set is acquired from the pre-constructed employee database, missing values may be formed in the acquired employee data under the conditions that the data cannot be acquired at the current time or the data is lost, so that the data in the characteristic data set is incomplete.

In detail, the missing value filling of the feature data set to obtain a standard feature data set includes:

searching a missing value data label of each employee data in the characteristic data set according to the characteristic label of the characteristic data set;

In the embodiment of the invention, the data label of the missing value is a characteristic label missing in the employee data, the characteristic data of other employees corresponding to the missing value data label is extracted from the characteristic data set to obtain the missing data set, and the missing value of each employee data is calculated based on the missing data set, so that the data really worth of is closer to the true value.

In the embodiment of the invention, all the feature tags in the feature data set can be searched in the feature data of each employee, and if the feature tags cannot be searched, the corresponding feature tags are the missing value feature tags.

Specifically, in the embodiment of the invention, the missing value in each employee data can be calculated by using a gray prediction method, an average filling method, a K nearest neighbor method and the like, so that the data of the feature data set after the missing value filling is complete and is closer to a true value, and the accuracy of predicting the subsequent employee risk behaviors is further improved.

In the embodiment of the present invention, the data in the static feature data set is feature data that does not change in a short time, such as information of a study, personality, address, and title, and the dynamic feature data set is feature data that changes constantly, such as attendance data, leave-on data, overtime data, mail volume data, and the like.

S3, constructing an initial prediction model by using the static characteristic data set, and calculating the initial job leaving probability of the characteristic data set by using the initial prediction model to obtain an initial probability set;

in the embodiment of the invention, the initial prediction model can be obtained by utilizing XGboost (Extreme Gradient Boosting) training, wherein the XGboost is an optimization algorithm based on a Gradient Boosting decision tree algorithm.

In detail, referring to fig. 2, the constructing an initial prediction model by using the static feature data set includes:

s21, carrying out normalization processing on the static characteristic data set;

s22, performing model training on a pre-constructed XGboost algorithm model based on the static feature data set after normalization processing to obtain an output probability value;

s23, calculating an optimal objective function of the XGboost algorithm model according to the output probability value, and determining an optimal decision tree in the XGboost algorithm model according to the optimal objective function;

and S24, determining the optimal parameters of the XGboost algorithm model according to the optimal decision tree to obtain an initial prediction model.

Specifically, the embodiment of the invention can map all the feature data in the same data scale by using methods such as the most value normalization, the mean variance normalization and the like, so that the result of model training is more accurate.

Further, the calculating an optimal objective function of the XGBoost algorithm model according to the output probability value includes:

performing second-order Taylor expansion on the target function to obtain a Taylor expansion formula, and removing a constant term in the Taylor expansion formula to obtain a function expansion formula;

In the embodiment of the invention, the XGboost algorithm model comprises a plurality of decision trees, each decision tree comprises different decision functions, the optimal decision function of each decision tree can be obtained through the optimal objective function, and the initial prediction model can be obtained by taking the optimal decision function as the optimal parameter of the XGboost algorithm model.

In the embodiment of the invention, the initial prediction model is constructed by utilizing the static characteristic data set, so that the initial job leaving prediction can be carried out on the staff, for example, the staff cannot have the idea of job leaving after six months or one year, and the characteristic data is provided for the construction of the subsequent staff job leaving prediction model.

S4, extracting time sequence characteristics in the dynamic characteristic data set by using a preset time sequence characteristic model to obtain a time sequence characteristic set;

in an embodiment of the present invention, the preset time sequence characteristic model is a TCN (Temporal Convolutional Network), and is composed of one-dimensional Convolutional layers having the same input and output

In detail, referring to fig. 3, the extracting, by using a preset time series feature model, the time series feature in the dynamic feature data set to obtain a time series feature set includes:

s31, performing convolution on the dynamic feature data by utilizing the convolution layer in the time sequence feature model to obtain a convolution feature map, and performing space pooling on the convolution feature map;

s32, performing convolution channel compression on the convolution characteristic diagram after the space pooling by using the convolution core in the convolution layer to obtain a compressed characteristic diagram;

and S33, performing expansion convolution operation on the compressed feature map through the convolution layer to obtain the time sequence feature of the feature data set.

In an optional embodiment of the present invention, the preset time sequence feature model may be obtained by training a large amount of staff dynamic feature data, the preset time sequence feature model includes a plurality of convolution layers, each convolution layer is composed of expanded one-dimensional convolution layers having the same input and output lengths, and the number of layers of the convolution layers and the size of the convolution core may be set differently according to actual needs to adjust the size of the output convolution feature map, so that the mined time sequence feature is more accurate.

Further, the performing, by the convolutional layer, a dilation convolution operation on the compressed feature map includes:

wherein K is the number of convolution layers, f _k Is the kth convolutional layer, X is the expansion convolutional sequence, X _t For the t-th dilation convolution operation, d is a dilation parameter,

is the x _t And (4) calculating the time sequence characteristics by the dilation convolution. .

In the embodiment of the invention, the expansion parameter is determined according to the number of intervals in the characteristic diagram, and the range of the next convolution calculation can be enlarged through the expansion coefficient, so that each convolution contains more characteristic information, and the obtained characteristic information contained in the time sequence characteristic is more comprehensive.

In the embodiment of the invention, the preset time sequence feature model is used for carrying out feature mining on the historical feature information and the feature change rule in the dynamic feature set, and the time sequence feature is extracted to further discover the recent behavior feature change of the staff and improve the accuracy of staff job leaving prediction.

And S5, constructing an employee prediction model according to the time sequence feature set and the initial probability set, and calculating the employee job leaving probability by using the job leaving prediction model.

In the embodiment of the invention, the risk behavior prediction model is obtained by utilizing the XGBoost algorithm training based on the timing characteristic set and the initial probability set, and the initial prediction probability in the initial probability set is used as a characteristic for training, so as to increase the comprehensiveness of the characteristic.

According to the embodiment of the invention, the staff job leaving prediction model is constructed through the time sequence characteristics and the prediction probability of the initial prediction model, the prediction model is established through distribution, the characteristic data is subdivided, and the static characteristic information and the dynamic characteristic information are fully utilized, so that the accuracy of the prediction of the job leaving prediction model is further improved.

Specifically, in the embodiment of the present invention, the probability of employee leaving is predicted by using the leave prediction model that is completed through pre-training, a leave risk alarm is generated when the probability is greater than or equal to a preset probability threshold, and a corresponding handler is timely reminded to intervene in advance to prevent sudden leave of the employee, for example, the probability threshold may be 70%, that is, when the leave prediction model is used to predict the leave of the feature data set of the target employee, the leave probability of the target employee is 75%, the leave risk of the target employee is higher, and the target employee needs to intervene in advance to prevent sudden leave of the target employee.

In an optional embodiment of the invention, after the employee resignation prediction model is obtained, the risk behavior prediction model may be encapsulated by using a flash microservice framework to deploy an employee resignation prediction system, and a flash _ unscheduler (timed task framework) timer may be set to perform early warning on employee resignation risk behaviors at regular time, so as to prevent risks such as enterprise business stagnation or confidential information leakage caused by sudden employee resignation behaviors.

The embodiment of the invention further acquires the characteristic data by acquiring the characteristic label in the working scene, so that the characteristic data comprises the characteristic data specific to the working scene, the characteristic data is more comprehensive, and the prediction accuracy of the risk behavior prediction model is favorably improved; and filling missing values in the feature data to obtain a standard feature data set, wherein the feature data set comprises a static feature data set and a dynamic feature data set. Subdividing the characteristic data set and modeling step by step to improve the prediction accuracy of the employee job leaving prediction model; constructing an initial prediction model by using the static characteristic data set to carry out risk prediction on the characteristic data set to obtain an initial probability set; extracting time sequence characteristics in the dynamic characteristic data set, and further enlarging the range of the characteristics; and constructing an employee job leaving prediction model by utilizing the time sequence characteristic set and the initial probability set, and calculating the employee job leaving probability by utilizing the job leaving prediction model so as to improve the prediction accuracy of employee job leaving prediction. Therefore, the staff leave prediction method provided by the invention can solve the problem of low accuracy in staff leave prediction.

Fig. 4 is a functional block diagram of an employee departure prediction apparatus according to an embodiment of the present invention.

The employee departure prediction apparatus 100 according to the present invention may be installed in an electronic device. According to the realized functions, the employee departure prediction apparatus 100 may include a feature data acquisition module 101, a missing value filling module 102, an initial prediction module 103, a time sequence feature extraction module 104, and an employee departure prediction model construction module 105. The module of the present invention, which may also be referred to as a unit, refers to a series of computer program segments that can be executed by a processor of an electronic device and can perform a fixed function, and are stored in a memory of the electronic device.

In the present embodiment, the functions regarding the respective modules/units are as follows:

the characteristic data acquisition module 101 is configured to acquire a characteristic label in a working scene, and acquire a characteristic data set in the working scene according to the characteristic label;

the missing value filling module 102 is configured to perform missing value filling on the feature data set to obtain a standard feature data set, where the feature data set includes a static feature data set and a dynamic feature data set;

the initial prediction module 103 is configured to construct an initial prediction model by using the static feature data set, and calculate an initial job leaving probability of the feature data set by using the initial prediction model to obtain an initial probability set;

the time sequence feature extraction module 104 is configured to extract time sequence features in the dynamic feature data set by using a preset time sequence feature model to obtain a time sequence feature set;

the employee job leaving prediction model building module 105 is configured to build an employee job leaving prediction model according to the time-series feature set and the initial probability set, and calculate an employee job leaving probability by using the job leaving prediction model.

In detail, in the embodiment of the present invention, when the modules described in the employee resignation prediction apparatus 100 are used, the same technical means as the employee resignation prediction method described in fig. 1 to fig. 3 are adopted, and the same technical effects can be produced, which are not described herein again.

Fig. 5 is a schematic structural diagram of an electronic device for implementing an employee attendance prediction method according to an embodiment of the present invention.

The electronic device 1 may comprise a processor 10, a memory 11, a communication bus 12 and a communication interface 13, and may further comprise a computer program, such as an employee attendance prediction program, stored in the memory 11 and executable on the processor 10.

In some embodiments, the processor 10 may be composed of an integrated circuit, for example, a single packaged integrated circuit, or may be composed of a plurality of integrated circuits packaged with the same function or different functions, and includes one or more Central Processing Units (CPUs), a microprocessor, a digital Processing chip, a graphics processor, a combination of various control chips, and the like. The processor 10 is a Control Unit (Control Unit) of the electronic device, connects various components of the whole electronic device by using various interfaces and lines, and executes various functions and processes data of the electronic device by running or executing programs or modules (for example, executing employee attendance prediction programs and the like) stored in the memory 11 and calling data stored in the memory 11.

The memory 11 includes at least one type of readable storage medium including flash memory, removable hard disks, multimedia cards, card-type memory (e.g., SD or DX memory, etc.), magnetic memory, magnetic disks, optical disks, etc. The memory 11 may in some embodiments be an internal storage unit of the electronic device, for example a removable hard disk of the electronic device. The memory 11 may also be an external storage device of the electronic device in other embodiments, such as a plug-in mobile hard disk, a Smart Media Card (SMC), a Secure Digital (SD) Card, a Flash memory Card (Flash Card), and the like, provided on the electronic device. Further, the memory 11 may also include both an internal storage unit and an external storage device of the electronic device. The memory 11 may be used not only to store application software installed in the electronic device and various types of data, such as codes of employee departure prediction programs, etc., but also to temporarily store data that has been output or is to be output.

The communication bus 12 may be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus. The bus may be divided into an address bus, a data bus, a control bus, etc. The bus is arranged to enable connection communication between the memory 11 and at least one processor 10 or the like.

The communication interface 13 is used for communication between the electronic device and other devices, and includes a network interface and a user interface. Optionally, the network interface may include a wired interface and/or a wireless interface (e.g., WI-FI interface, bluetooth interface, etc.), which are typically used to establish a communication connection between the electronic device and other electronic devices. The user interface may be a Display (Display), an input unit, such as a Keyboard (Keyboard), and optionally a standard wired interface, a wireless interface. Alternatively, in some embodiments, the display may be an LED display, a liquid crystal display, a touch-sensitive liquid crystal display, an OLED (Organic Light-Emitting Diode) touch device, or the like. The display, which may also be referred to as a display screen or display unit, is suitable, among other things, for displaying information processed in the electronic device and for displaying a visualized user interface.

Only electronic devices having components are shown, and those skilled in the art will appreciate that the structures shown in the figures do not constitute limitations on the electronic devices, and may include fewer or more components than shown, or some components in combination, or a different arrangement of components.

For example, although not shown, the electronic device may further include a power supply (such as a battery) for supplying power to each component, and preferably, the power supply may be logically connected to the at least one processor 10 through a power management device, so that functions of charge management, discharge management, power consumption management and the like are realized through the power management device. The power supply may also include any component of one or more dc or ac power sources, recharging devices, power failure detection circuitry, power converters or inverters, power status indicators, and the like. The electronic device may further include various sensors, a bluetooth module, a Wi-Fi module, and the like, which are not described herein again.

It is to be understood that the described embodiments are for purposes of illustration only and that the scope of the appended claims is not limited to such structures.

The employee attendance prediction program stored in the memory 11 of the electronic device 1 is a combination of instructions that, when executed in the processor 10, enable:

acquiring feature labels of enterprise employees in different fields in an enterprise in a working scene, and acquiring a feature data set in the working scene according to the feature labels;

filling missing values in the feature data set to obtain a standard feature data set, wherein the feature data set comprises a static feature data set and a dynamic feature data set;

Specifically, the specific implementation method of the instruction by the processor 10 may refer to the description of the relevant steps in the embodiment corresponding to the drawings, which is not described herein again.

Further, the integrated modules/units of the electronic device 1, if implemented in the form of software functional units and sold or used as separate products, may be stored in a computer readable storage medium. The computer readable storage medium may be volatile or non-volatile. For example, the computer-readable medium may include: any entity or device capable of carrying said computer program code, a recording medium, a usb-disk, a removable hard disk, a magnetic diskette, an optical disk, a computer Memory, a Read-Only Memory (ROM).

The present invention also provides a computer-readable storage medium, storing a computer program which, when executed by a processor of an electronic device, may implement:

In the embodiments provided in the present invention, it should be understood that the disclosed apparatus, device and method can be implemented in other ways. For example, the above-described apparatus embodiments are merely illustrative, and for example, the division of the modules is only one logical functional division, and other divisions may be realized in practice.

The modules described as separate parts may or may not be physically separate, and parts displayed as modules may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of the present embodiment.

In addition, functional modules in the embodiments of the present invention may be integrated into one processing unit, or each unit may exist alone physically, or two or more units are integrated into one unit. The integrated unit can be realized in a form of hardware, or in a form of hardware plus a software functional module.

It will be evident to those skilled in the art that the invention is not limited to the details of the foregoing illustrative embodiments, and that the present invention may be embodied in other specific forms without departing from the spirit or essential attributes thereof.

The present embodiments are therefore to be considered in all respects as illustrative and not restrictive, the scope of the invention being indicated by the appended claims rather than by the foregoing description, and all changes which come within the meaning and range of equivalency of the claims are therefore intended to be embraced therein. Any reference signs in the claims shall not be construed as limiting the claim concerned.

The block chain is a novel application mode of computer technologies such as distributed data storage, point-to-point transmission, a consensus mechanism, an encryption algorithm and the like. A block chain (Blockchain), which is essentially a decentralized database, is a series of data blocks associated by using a cryptographic method, and each data block contains information of a batch of network transactions, so as to verify the validity (anti-counterfeiting) of the information and generate a next block. The blockchain may include a blockchain underlying platform, a platform product service layer, an application service layer, and the like.

The embodiment of the application can acquire and process related data based on an artificial intelligence technology. Among them, Artificial Intelligence (AI) is a theory, method, technique and application system that simulates, extends and expands human Intelligence using a digital computer or a machine controlled by a digital computer, senses the environment, acquires knowledge and uses the knowledge to obtain the best result.

Furthermore, it will be obvious that the term "comprising" does not exclude other elements or steps, and the singular does not exclude the plural. A plurality of units or means recited in the system claims may also be implemented by one unit or means in software or hardware. The terms first, second, etc. are used to denote names, but not any particular order.

Finally, it should be noted that the above embodiments are only intended to illustrate the technical solutions of the present invention and not to limit the same, and although the present invention is described in detail with reference to the preferred embodiments, it should be understood by those skilled in the art that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A method for predicting employee job leaving, the method comprising:

constructing an initial prediction model by using the static characteristic data set, and calculating the initial job leaving probability of the characteristic data set by using the initial prediction model to obtain an initial probability set;

extracting the time sequence characteristics of the dynamic characteristic data set by using a preset time sequence characteristic model to obtain a time sequence characteristic set;

2. The employee attendance prediction method according to claim 1, wherein the obtaining of the feature data set in the work scenario from the feature tag comprises:

3. The employee attendance prediction method according to claim 1, wherein said missing value padding of the feature data set to obtain a standard feature data set comprises:

4. The employee attendance prediction method of claim 1 wherein said building an initial prediction model using said static feature data set comprises:

carrying out normalization processing on the static characteristic data set;

model training is carried out on a pre-constructed XGboost algorithm model based on the static characteristic data set after normalization processing, and an output probability value is obtained;

5. The employee departure prediction method according to claim 4, wherein said calculating an optimal objective function of said XGboost algorithm model based on said output probability values comprises:

6. The employee attendance prediction method according to any one of claims 1 to 5, wherein the extracting time series features in the dynamic feature data set using a preset time series feature model to obtain a time series feature set comprises:

and performing expansion convolution operation on the compressed feature map through the convolution layer to obtain the time sequence feature of the feature data set.

7. The employee resignation prediction method of claim 6, wherein said performing a dilation convolution operation on said compressed feature map with said convolution layer comprises:

is the x _t And (4) calculating the obtained time sequence characteristics by using the dilation convolution.

8. An employee departure prediction apparatus, comprising:

and the employee departure prediction model construction module is used for constructing an employee departure prediction model according to the time sequence feature set and the initial probability set, and calculating employee departure probability by utilizing the departure prediction model.

9. An electronic device, characterized in that the electronic device comprises:

at least one processor; and the number of the first and second groups,

a memory communicatively coupled to the at least one processor; wherein, the first and the second end of the pipe are connected with each other,

the memory stores a computer program executable by the at least one processor to enable the at least one processor to perform the employee job leaving prediction method of any one of claims 1 to 7.

10. A computer-readable storage medium storing a computer program, wherein the computer program, when executed by a processor, implements the employee departure prediction method according to any one of claims 1 to 7.