WO2020042584A1 - 非绩优人员培训方法、系统、计算机装置及存储介质 - Google Patents

非绩优人员培训方法、系统、计算机装置及存储介质 Download PDF

Info

Publication number
WO2020042584A1
WO2020042584A1 PCT/CN2019/077511 CN2019077511W WO2020042584A1 WO 2020042584 A1 WO2020042584 A1 WO 2020042584A1 CN 2019077511 W CN2019077511 W CN 2019077511W WO 2020042584 A1 WO2020042584 A1 WO 2020042584A1
Authority
WO
WIPO (PCT)
Prior art keywords
factor
performance
personnel
behavioral
value
Prior art date
Application number
PCT/CN2019/077511
Other languages
English (en)
French (fr)
Inventor
邓悦
金戈
徐亮
肖京
Original Assignee
平安科技（深圳）有限公司
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 平安科技（深圳）有限公司 filed Critical 平安科技（深圳）有限公司
Publication of WO2020042584A1 publication Critical patent/WO2020042584A1/zh

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
    • G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/08—Learning methods
    • 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
    • G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
    • G06Q50/10—Services
    • G06Q50/20—Education
    • G06Q50/205—Education administration or guidance
    • G06Q50/2057—Career enhancement or continuing education service

Definitions

  • the present application relates to the field of data processing, and in particular, to a method, system, computer device, and storage medium for training non-excellent personnel.
  • the present application provides a method, system, computer device, and storage medium for training non-excellent personnel, which can implement targeted training scheme generation according to different training personnel.
  • An embodiment of the present application provides a method for training non-outstanding personnel, the method includes: obtaining sample data of a plurality of excellent personnel, wherein the sample data includes behavior factor data and non-behavioral factor data, and Belonging to multiple types of excellence, each of the high performing personnel is classified as one type of excellence;
  • An embodiment of the present application provides a system for training non-excellent personnel.
  • the system includes:
  • An acquisition module for obtaining sample data of a plurality of high-performing persons, wherein the sample data includes behavioral factor data and non-behavioral factor data, a plurality of the high-performing persons are classified into multiple types of high-performing persons, and each of the high-performing persons is classified as belonging to A type of merit
  • a first calculation module configured to input non-behavioral factor data of non-outstanding personnel into the potential performance identification model and calculate a probability value that the non-outstanding personnel grows into each type of performance;
  • a selection module configured to select a maximum probability value from all the calculated probability values, and use the type of merit corresponding to the maximum probability value as the potential merit type of the non-excellent person;
  • a second calculation module configured to calculate a mean value and a standard deviation of each behavior factor of a high-performance person in the potential performance type, and compare each behavior factor value of the non-high-performance person with each of the potential performance type Make a one-to-one comparison of the mean of the behavioral factors;
  • a judging module configured to determine that when a difference between a behavior factor value of the non-outstanding person and a corresponding average value of the behavior factor is greater than a preset multiple of a standard deviation, the non-outstanding person needs to select a training course for improving the behavior factor .
  • An embodiment of the present application provides a computer device.
  • the computer device includes a processor and a memory.
  • the memory stores a plurality of computer-readable instructions.
  • the processor is configured to execute the computer-readable instructions stored in the memory, such as The steps of the training method for non-high-performing personnel described earlier.
  • An embodiment of the present application provides a non-volatile readable storage medium on which computer-readable instructions are stored.
  • the steps of the method for training non-high performers as described above are implemented. .
  • the above non-outstanding personnel training method, system, computer device, and non-volatile readable storage medium are based on machine learning and non-behavioral factor data of outperforming personnel to establish and train a potential excellent performance identification model, and use this model to identify non-outstanding personnel.
  • the type of merit of the potential for growth of high-performing personnel A behavioral factor of non-high-performing personnel is compared with the average of the high-performing person's behavior factor to determine whether a non-high-performing person needs to choose a training course that enhances the behavior factor to achieve objective determination of non-performing personnel.
  • the type of performance that high-performing personnel are most likely to grow, and then from the perspective of the day after tomorrow, find the shortcomings of non-high-performing personnel and the target performance type, and improve the effectiveness of personnel training.
  • FIG. 1 is a flowchart of steps of a method for training non-outstanding personnel in an embodiment of the present application.
  • FIG. 2 is a flowchart of steps of a method for training non-excellent personnel in another embodiment of the present application.
  • FIG. 3 is a functional module diagram of a non-outstanding personnel training system in an embodiment of the present application.
  • FIG. 4 is a schematic diagram of a computer device according to an embodiment of the present application.
  • the non-excellent personnel training method of the present application is applied in one or more computer devices.
  • the computer device is a device capable of automatically performing numerical calculations and / or information processing in accordance with instructions set or stored in advance.
  • the hardware includes, but is not limited to, a microprocessor and an Application Specific Integrated Circuit (ASIC). , Programmable Gate Array (Field-Programmable Gate Array, FPGA), Digital Processor (Digital Signal Processor, DSP), Embedded Equipment, etc.
  • ASIC Application Specific Integrated Circuit
  • FIG. 1 is a flowchart of steps in a preferred embodiment of a method for training non-excellent personnel of the present application. According to different requirements, the order of the steps in the flowchart can be changed, and some steps can be omitted.
  • the method for training non-outstanding personnel includes the following steps.
  • Step S11 Obtain sample data of a plurality of high-performing persons, wherein the sample data includes behavioral factor data and non-behavioral factor data, a plurality of the high-performing persons are classified into multiple types of high-performing persons, and each of the high-performing persons is classified into one type Performance type.
  • the sample data of the top performers stored in the top performer sample database may be obtained by connecting to a top performer sample database through an access network.
  • the high-performance personnel sample database may collect sample data of multiple high-performance personnel by means of big data, or may receive sample data of multiple high-performance personnel entered by humans.
  • a plurality of the high-performance personnel belong to multiple types of high-performance, but each of the high-performance personnel belongs to only one high-performance type, that is, each of the high-performance personnel cannot belong to multiple high-performance types at the same time.
  • the multiple types of performance may include three types of performance, which are resource-based, learning-based, and hard-working.
  • the resource type can refer to people with strong business ability and working ability
  • the learning type can refer to people with strong learning ability and strong growth ability
  • the diligent type can refer to people with long learning time and long working hours per day.
  • the following uses the example of high-performing and non-high-performing personnel as employees of Company A as an example.
  • Company A includes 1,000 employees.
  • the 1,000 employees can be classified into two types: high-performing and non-high-performing.
  • Each employee can only be classified as high-performing or non-high-performing.
  • the division of non-high performers can be divided according to the division rules corresponding to multiple high performers.
  • each employee's learning ability and growth ability can be used to determine whether an employee is a high performer of learning type.
  • resource-based performance excellence you can determine whether an employee belongs to resource-based performance excellence according to the business capabilities of each employee.
  • the types of top performers include resource, learning, and hard work.
  • 200 of them were classified as high-performing personnel, and the remaining 800 employees were classified as non-high-performing personnel.
  • each high-performing person belongs to a type of high-performing
  • resource-based high-performing personnel includes 80
  • learning high-performing personnel includes 50
  • hard-working high-performing personnel includes 70.
  • the sample data of the top performers and the non-top performers both include behavioral factor data and non-behavioral factor data.
  • the behavior factor data includes multiple behavior factors
  • the non-behavior factor data includes multiple non-behavior factors
  • each of the behavior factors corresponds to a behavior factor value
  • each of the non-behavior factors corresponds to a non-behavior factor value
  • the high-performing person and the non-high-performing person have the same behavior factor and non-behavior factor.
  • the behavior factor preferably refers to an acquired behavior factor or a behavior factor that can be changed and cultivated in a short period of time.
  • the behavior factor includes a range of activities, frequency of entering and exiting various places, attendance, APP activity, communication ability, and the like.
  • the non-behavioral factors preferably refer to innate behavioral factors or behavioral factors that are difficult to change and cultivate in a short time, for example, the non-behavioral factors include age, gender, education, and the like.
  • a non-behavioral factor is age
  • the value of the corresponding non-behavioral factor is age. If the age of the top performer A1 is 27 and the age of the top performer A2 is 30, then the value of the age of the top performer A1 is the non-behavioral factor. It is 27, and the non-behavioral factor value of the age of the top performer A2 is 30; if a behavioral factor is communication ability, the behavioral factor value of the behavioral factor can be scored according to the actual communication ability level of the employee (can be based on a preset scoring standard) Get the corresponding behavior factor value. In step S11, obtaining the sample data of a plurality of high-performing persons is to obtain behavioral factor data and non-behavioral factor data of a plurality of high-performing persons.
  • Step S12 Establish and train a potential performance recognition model based on the non-behavioral factor data of a plurality of performance performers.
  • the potential performance recognition model may be a classification model trained based on a neural network model and non-behavioral factor data of a plurality of performance excellence personnel. Specifically, a neural network model may be established first, where the neural network model includes an input layer, multiple hidden layers, and an output layer, and then the neural network model is obtained by training a plurality of non-behavioral factor data of the performers. The potential performance recognition model.
  • the input layer of the neural network model is used to receive multiple non-behavioral factor data of the top performers, each hidden layer includes multiple nodes (neurons), and each node in each hidden layer is configured to pair The output of at least one node of an adjacent lower layer in the model performs a linear or non-linear transformation.
  • the input of the nodes in the upper hidden layer can be based on the output of one node or several nodes in the adjacent lower layer, and each hidden layer has a corresponding weight, which is obtained based on the training sample data.
  • the model can be trained by using a supervised learning process to obtain the initial weights of each hidden layer.
  • the backpropagation (BP) algorithm can be used to adjust the weight of each hidden layer.
  • the output layer of the neural network model is used to receive the output signal from the last hidden layer.
  • step S12 may specifically include:
  • model prediction accuracy rate is not less than the preset threshold, use the neural network model that has been trained as the potential performance recognition model.
  • model prediction accuracy rate is less than the preset threshold, adjust the parameters of the neural network model, and use the training set to retrain the adjusted neural network model until the model prediction obtained by the validation set is verified
  • accuracy rate is not less than the preset threshold, wherein the parameters of the neural network model include the total number of layers, the number of neurons in each layer, and the like.
  • the parameter for adjusting the neural network model may be adjusting the total number of layers of the neural network model and / or the number of neurons in each layer.
  • the training set is used to train a neural network model
  • the validation set is used to verify a trained neural network model.
  • the neural network model may be trained by using the training set to obtain an intermediate model, and then the non-behavioral factor data of the performers in the verification set may be input into the intermediate model to perform classification verification of the superiority type.
  • a model prediction accuracy rate can be statistically obtained; and then it is judged whether the intermediate model prediction accuracy rate is less than a preset threshold value; if the intermediate model prediction accuracy rate is not less than the preset threshold value, this intermediate model is indicated
  • the classification effect is good, and the use of the intermediate model can be directly used as the recognition model of potential performance. If the prediction accuracy rate of the model is less than the preset threshold, it indicates that the classification effect of the intermediate model is not good and needs to be performed.
  • the parameters of the neural network model can be adjusted, and the adjusted neural network model is retrained using the training set to obtain a new intermediate model, and then the verified set is used again to retrieve the intermediate model.
  • the retrieved intermediate model can be used as the potential. Excellent performance recognition model; if the prediction accuracy of the new model is still less than the preset threshold, the above steps need to be repeated again until the model prediction accuracy obtained through the verification set is not less than the preset threshold.
  • the preset threshold may be set according to actual usage requirements. For example, the preset threshold is set to 95%, that is, the model prediction accuracy rate needs to be not less than 95%.
  • Step S13 Input the non-behavioral factor data of the non-outstanding personnel into the potential performance recognition model and calculate the probability value of the non-outstanding personnel to grow into each type of outstanding performance.
  • the potential performance recognition model can identify potential performance excellence types of non-high-performing personnel.
  • the non-behavioral factor data of non-high-performance personnel can be used as the potential high-performance
  • the input of the recognition model, and the output of the potential performance recognition model is regarded as the probability of non-high-performing people growing into each type of performance.
  • each probability value represents a probability value that the non-outstanding person grows into an outstanding person of each excellent type.
  • a maximum probability value may be selected from all calculated probability values, and The type of merit corresponding to the maximum probability value is used as the potential merit type of the non-outstanding person, so as to identify the most potential merit type of growth of the non-outstanding person.
  • the probability that non-outstanding personnel A1 will grow into resource-based high-performing personnel is 0.6
  • the probability of growing into learning-based high-performing personnel is 0.7
  • the probability of growing into hard-working high-performing personnel is 0.55.
  • the probability value of growing into a learning excellence person is the largest, and the learning type is judged as the potential excellence type of the non-outstanding person A1.
  • a type of merit corresponding to a maximum probability value may be randomly selected as a potential type of merit of the non-excellent person.
  • the probability of non-outperformer A2 growing into a resource excellent performer is 0.8
  • the probability of growing into a learning excellent performer is 0.7
  • the probability of growing into a hardworking outstanding performer is 0.8.
  • the probability value of hard-working high-performance personnel and growing into resource-based high-performance personnel are both 0.8, you can choose to judge hard-working type as the potential high-performance type of the non-high-performance person A2, or you can determine resource-based high-performance person A2 as the Potential development performance type.
  • a type of merit corresponding to a maximum probability value that matches a preset requirement may also be selected as the potential type of merit of the non-highly-performing person.
  • the preset demand may be a company's current critical staffing type requirements. If the company has the largest demand for resource-based high-performance personnel, the demand for learning-based high-performance personnel is the second, and the demand for hard-working high-performance personnel is the lowest.
  • the probability of growing into a resource high performing person is 0.8
  • the probability of growing into a learning high performing person is 0.7
  • the probability of growing into a hardworking high performing person is 0.8.
  • the probability value is 0.8 and the company has the greatest demand for resource-based high-performance personnel, then the resource type can be judged as the potential high-performance type of the non-high-performance personnel A2.
  • step S14 may be used to determine a potential high-performance type of a non-high-performance person.
  • the potential high-performance type includes data of behavior factors of multiple high-performance personnel. By calculating each behavior factor of the high-performance person in the high-performance potential type, Mean value and standard deviation, and then perform a one-to-one correspondence comparison between each behavior factor value of the non-high performing person and the average value of each behavior factor of the potential high performing type to realize judging each behavior factor of the low performing person Whether it needs to be promoted.
  • each behavior factor of the non-high performing person has been quantified in advance, and each behavior factor has a behavior factor value.
  • each behavior factor of each outstanding person has been quantified in advance, that is, each behavior factor of each outstanding person has a behavior factor value.
  • each person (excellent and non-excellent) has 50 behavioral factors
  • the potential performance type of non-excellent person A1 obtained in step S14 is resource type.
  • the average behavioral factor of the behavioral factor a1 in the 100 top performers is 70
  • the standard deviation is 4
  • the behavioral factor a2 is calculated in the 100
  • the average performance factor of the top performers was 72 and the standard deviation was 3.5.
  • the value of the behavior factor a1 of the non-outstanding person is compared with the average value 70
  • the value of the behavior factor a2 of the non-outstanding person is compared with the average value 72.
  • Step S16 When the difference between a behavior factor value of the non-highly-performing person and a corresponding average value of the behavior factor is greater than a preset multiple of a standard deviation, it is determined that the non-high-performing person needs to select a training course for improving the behavior factor.
  • the preset multiple may be set according to an actual usage scenario, for example, set to a standard deviation of 2 times.
  • the standard deviation may be set to 3 times or 4 times.
  • an example is taken with a standard deviation of 2 times, and the value of the behavior factor a1 of the non-outstanding person is compared with the corresponding average value of 70 to determine whether the difference is greater than the standard deviation of 2 times, that is, the non- Whether the difference between the value of the behavior factor a1 of the excellent performer and the corresponding mean 70 is greater than 8 (2 * 4), and if the difference is greater than 8, it is judged that the non-high performer needs to choose a training course to improve the behavior factor a1.
  • the non-outstanding personnel training method shown in FIG. 2 further includes step S17.
  • step S17 when the difference between a behavior factor value of the non-high-performing person and the corresponding average value of the behavior factor is not greater than a preset multiple of the standard deviation, it is determined that the non-high-performing person does not need to select a training course for improving the behavior factor. .
  • the standard deviation of 2 times is also used as an example for illustration.
  • the value of the behavior factor a1 of the non-outstanding person is compared with the corresponding average value 70 to determine whether the difference is greater than the standard deviation of 2 times. Whether the difference between the value of the behavior factor a1 of the non-outstanding person and the corresponding mean 70 is greater than 8 (2 * 4). If the difference is not greater than 8, it is judged that the non-outstanding person does not need to choose a training course to improve the behavior factor a1 .
  • the value of the behavior factor a1 of a non-outstanding person A1 is 63, and the value of the behavior factor a2 is 60.
  • FIG. 3 is a functional module diagram of a preferred embodiment of a non-outstanding personnel training system of the present application.
  • the non-excellent personnel training system 10 may include an acquisition module 101, a establishment module 102, a first calculation module 103, a selection module 104, a second calculation module 105, and a determination module 106.
  • the obtaining module 101 is configured to obtain sample data of a plurality of high-performance personnel, where the sample data includes behavior factor data and non-behavioral factor data, a plurality of the high-performance personnel are classified into multiple types of high-performance personnel, and each of the high-performance personnel is classified as belonging to A type of merit.
  • the obtaining module 101 may be connected to a sample database of high performing personnel through an access network, and then obtain sample data of high performing personnel stored in the sample database of high performing personnel.
  • the sample database of high-performing personnel may collect sample data of multiple high-performing personnel by means of big data, or may receive sample data of multiple high-performing personnel entered manually.
  • a plurality of the high-performance personnel belong to multiple types of high-performance, but each of the high-performance personnel belongs to only one high-performance type, that is, each of the high-performance personnel cannot belong to multiple high-performance types at the same time.
  • the multiple types of performance may include three types of performance, which are resource-based, learning-based, and hard-working.
  • the resource type can refer to people with strong business ability and working ability
  • the learning type can refer to people with strong learning ability and strong growth ability
  • the diligent type can refer to people with long learning time and long working hours per day.
  • the following uses the example of high-performing and non-high-performing personnel as employees of Company A as an example.
  • Company A includes 1,000 employees.
  • the 1,000 employees can be classified into two types: high-performing and non-high-performing.
  • Each employee can only be classified as high-performing or non-high-performing.
  • the division of non-high performers can be divided according to the division rules corresponding to multiple high performers.
  • each employee's learning ability and growth ability can be used to determine whether an employee is a high performer of learning type.
  • resource-based performance excellence you can determine whether an employee belongs to resource-based performance excellence according to the business capabilities of each employee.
  • the types of top performers include resource, learning, and hard work.
  • 200 of them were classified as high-performing personnel, and the remaining 800 employees were classified as non-high-performing personnel.
  • each high-performing person belongs to a type of high-performing
  • resource-based high-performing personnel includes 80
  • learning high-performing personnel includes 50
  • hard-working high-performing personnel includes 70.
  • the sample data of the top performers and the non-top performers both include behavioral factor data and non-behavioral factor data.
  • the behavior factor data includes multiple behavior factors
  • the non-behavior factor data includes multiple non-behavior factors
  • each of the behavior factors corresponds to a behavior factor value
  • each of the non-behavior factors corresponds to a non-behavior factor value
  • the high-performing person and the non-high-performing person have the same behavior factor and non-behavior factor.
  • the behavior factor preferably refers to an acquired behavior factor or a behavior factor that can be changed and cultivated in a short period of time.
  • the behavior factor includes a range of activities, frequency of entering and exiting various places, attendance, APP activity, communication ability, and the like.
  • the non-behavioral factors preferably refer to innate behavioral factors or behavioral factors that are difficult to change and cultivate in a short time, for example, the non-behavioral factors include age, gender, education, and the like.
  • a non-behavioral factor is age
  • the value of the corresponding non-behavioral factor is age. If the age of the top performer A1 is 27 and the age of the top performer A2 is 30, then the value of the age of the top performer A1 is the non-behavioral factor. It is 27, and the non-behavioral factor value of the age of the top performer A2 is 30; if a behavioral factor is communication ability, the behavioral factor value of the behavioral factor can be scored according to the actual communication ability level of the employee (can be based on a preset scoring standard) Get the corresponding behavior factor value. In step S11, obtaining the sample data of a plurality of high-performing persons is to obtain behavioral factor data and non-behavioral factor data of a plurality of high-performing persons.
  • the establishing module 102 is configured to establish and train a potential high performance recognition model according to the non-behavioral factor data of a plurality of high performance personnel.
  • the potential performance recognition model may be a classification model trained based on a neural network model and non-behavioral factor data of a plurality of performance excellence personnel.
  • the establishment module 102 may first establish a neural network model including an input layer, a plurality of hidden layers, and an output layer, and then use a plurality of non-behavioral factor data of the outstanding person to perform a neural network analysis on the neural network.
  • the network model is trained to obtain the potential performance recognition model.
  • the input layer of the neural network model is used to receive multiple non-behavioral factor data of the top performers, each hidden layer includes multiple nodes (neurons), and each node in each hidden layer is configured to pair The output of at least one node of an adjacent lower layer in the model performs a linear or non-linear transformation.
  • the input of the nodes in the upper hidden layer can be based on the output of one node or several nodes in the adjacent lower layer, and each hidden layer has a corresponding weight, which is obtained based on the training sample data.
  • the model can be trained by using a supervised learning process to obtain the initial weights of each hidden layer.
  • the backpropagation (BP) algorithm can be used to adjust the weight of each hidden layer.
  • the output layer of the neural network model is used to receive the output signal from the last hidden layer.
  • the manner in which the establishment module 102 establishes and trains a potential performance recognition model may specifically include:
  • model prediction accuracy rate is not less than the preset threshold, use the neural network model that has been trained as the potential performance recognition model.
  • model prediction accuracy rate is less than the preset threshold, adjust the parameters of the neural network model, and use the training set to retrain the adjusted neural network model until the model prediction obtained by the validation set is verified
  • accuracy rate is not less than the preset threshold, wherein the parameters of the neural network model include the total number of layers, the number of neurons in each layer, and the like.
  • the parameter for adjusting the neural network model may be adjusting the total number of layers of the neural network model and / or the number of neurons in each layer.
  • the training set is used to train a neural network model
  • the validation set is used to verify a trained neural network model.
  • the establishment module 102 may first use the training set to train a neural network model to obtain an intermediate model, and then input the non-behavioral factor data of the top performers in the verification set to the intermediate model. Classification and verification of superiority types, and a model prediction accuracy rate can be statistically obtained according to each verification result; and then determining whether the prediction accuracy rate of the intermediate model is less than a preset threshold; if the prediction accuracy rate of the intermediate model is not less than the preset threshold , Indicating that the intermediate model has a better classification effect and meets the needs of use, and the intermediate model can be directly used as the potential performance recognition model; if the model prediction accuracy rate is less than the preset threshold, the intermediate model classification effect is indicated Not good, needs to be improved.
  • the parameters of the neural network model can be adjusted, and the adjusted neural network model is re-trained using the training set to obtain a new intermediate model, and then the validation set is used again.
  • the retrieved intermediate model is verified to obtain a new model prediction accuracy rate, and then judged Determine whether the prediction accuracy of the new model is less than a preset threshold. If the prediction accuracy of the new model is not less than the preset threshold, it indicates that the retrieved intermediate model has a better classification effect and meets the needs of use.
  • the intermediate model is used as the potential performance recognition model; if the prediction accuracy of the new model is still less than the preset threshold, the above steps need to be repeated again until the model prediction accuracy obtained through the verification set is not less than the preset threshold.
  • the preset threshold may be set according to actual usage requirements. For example, the preset threshold is set to 95%, that is, the model prediction accuracy rate needs to be not less than 95%.
  • the first calculation module 103 is configured to input non-behavioral factor data of non-outstanding personnel into the potential performance recognition model and calculate a probability value that the non-outstanding personnel will grow into each type of superiority.
  • the potential merit recognition model can realize the identification of potential merit types of non-high performers.
  • the non-behavioral factor data of non-high performers can be used as The input of the potential performance recognition model is described, and the output of the potential performance recognition model is regarded as the probability that a non-high performance person grows into each performance type.
  • the selection module 104 is configured to select a maximum probability value from all the calculated probability values, and use the type of merit corresponding to the maximum probability value as the type of potential merit of the non-excellent person.
  • each probability value represents a probability value that the non-outstanding person grows into an outstanding person of each excellent type
  • the selection module 104 may select a maximum value from all the calculated probability values. Probability value, and using the type of merit corresponding to the maximum probability value as the potential merit type of the non-highly-performing person, so as to identify the most potential merit type of growth of the non-high-performance person.
  • the probability that non-outstanding personnel A1 will grow into resource-based high-performing personnel is 0.6
  • the probability of growing into learning-based high-performing personnel is 0.7
  • the probability of growing into hard-working high-performing personnel is 0.55.
  • the probability value of growing into a learning excellence person is the largest, and the learning type is judged as the potential excellence type of the non-outstanding person A1.
  • the selection module 104 may randomly select a type of merit corresponding to a maximum probability value as the potential type of merit of the non-outstanding person. For example, according to the potential performance recognition model, it is obtained that the probability of non-outperformer A2 growing into a resource excellent performer is 0.8, the probability of growing into a learning excellent performer is 0.7, and the probability of growing into a hardworking outstanding performer is 0.8.
  • the probability value of hard-working high-performance personnel and growing into resource-based high-performance personnel are both 0.8, you can choose to judge hard-working type as the potential high-performance type of the non-high-performance person A2, or you can determine resource-based high-performance person A2 as the Potential development performance type.
  • the selection module 104 may also select a type of merit corresponding to the maximum probability value that matches a preset requirement as the potential type of merit of the non-highly qualified person.
  • the preset demand may be a company's current critical staffing type requirements. If the company has the largest demand for resource-based high-performance personnel, the demand for learning-based high-performance personnel is the second, and the demand for hard-working high-performance personnel is the lowest. At this time, if the potential high performance recognition model is used to obtain the non-high performing person A2, the probability of growing into a resource high performing person is 0.8, the probability of growing into a learning high performing person is 0.7, and the probability of growing into a hardworking high performing person is 0.8.
  • the probability value is 0.8 and the company has the greatest demand for resource-based high-performance personnel, then the resource type can be judged as the potential high-performance type of the non-high-performance personnel A2.
  • the second calculation module 105 is configured to calculate a mean value and a standard deviation of each behavior factor of the high-performance person in the potential performance type, and compare each behavior factor value of the non-high-performance person with each of the potential performance type The mean values of behavioral factors are compared one-to-one.
  • the selection module 104 may be used to determine a potential performance type of a non-high performing person.
  • the potential performance type includes data of behavior factors of a plurality of high performing personnel.
  • the second calculation module 105 may be used to calculate the performance factor data. The mean value and standard deviation of each behavior factor of the high-performing person in the potential performance type, and then one-to-one correspondence between each behavior factor value of the non-high-performing person and the average value of each behavior factor of the potential performance type It is realized to judge whether each behavior factor of the non-outstanding person needs to be improved.
  • each behavior factor of the non-high performing person has been quantified in advance, and each behavior factor has a behavior factor value.
  • each behavior factor of each outstanding person has been quantified in advance, that is, each behavior factor of each outstanding person has a behavior factor value.
  • each person (excellent and non-excellent) has 50 behavioral factors
  • the potential performance type of non-excellent person A1 obtained in step S14 is resource type.
  • the average behavioral factor of the behavioral factor a1 in the 100 top performers is 70
  • the standard deviation is 4
  • the behavioral factor a2 is calculated in the 100
  • the average performance factor of the top performers was 72 and the standard deviation was 3.5.
  • the value of the behavior factor a1 of the non-outstanding person is compared with the average value 70
  • the value of the behavior factor a2 of the non-outstanding person is compared with the average value 72.
  • the judging module 106 is configured to determine that when a difference between a behavior factor value of the non-outstanding person and a corresponding average value of the behavior factor is greater than a preset multiple of a standard deviation, the non-outstanding person needs to select a training course for improving the behavior factor .
  • the preset multiple may be set according to an actual usage scenario, for example, set to a standard deviation of 2 times.
  • the standard deviation may be set to 3 times or 4 times.
  • an example is taken with a standard deviation of 2 times, and the value of the behavior factor a1 of the non-outstanding person is compared with the corresponding average value of 70 to determine whether the difference is greater than the standard deviation of 2 times, that is, the non- Whether the difference between the behavior factor a1 of the top performer and the corresponding mean 70 is greater than 8 (2 * 4); if the difference is greater than 8, the determination module 106 determines that the non-performer needs to choose to improve the behavior factor a1 Training Courses.
  • the determination module 106 determines that the non-outstanding person needs to select a training course for improving the behavior factor a2.
  • the determining module 106 is further configured to determine the non-meritorious person when the difference between a behavior factor value of the non-meritorious person and the corresponding mean of the behavior factor is not greater than a preset multiple of a standard deviation. The person does not need to choose a training course that promotes this behavior factor.
  • the standard deviation of 2 times is also used as an example for illustration.
  • the value of the behavior factor a1 of the non-outstanding person is compared with the corresponding average value 70 to determine whether the difference is greater than the standard deviation of 2 times. Whether the difference between the behavior factor a1 of the non-outstanding person and the corresponding mean 70 is greater than 8 (2 * 4); if the difference is not greater than 8, the determination module 106 determines that the non-outstanding person does not need to choose to improve the behavior Training course for factor a1.
  • the determination module 106 determines that the non-outstanding person does not need to select a training course for improving the behavior factor a2.
  • the value of the behavior factor a1 of a non-outstanding person A1 is 63, and the value of the behavior factor a2 is 60.
  • the determination module 106 determines the non-outstanding person A1 needs to choose a training course to improve the behavior factor a2.
  • FIG. 4 is a schematic diagram of a preferred embodiment of a computer device of the present application.
  • the computer device 1 includes a memory 20, a processor 30, and computer-readable instructions 40 stored in the memory 20 and executable on the processor 30, such as non-excellent personnel training programs.
  • the processor 30 executes the computer-readable instructions 40
  • the steps in the embodiment of the method for training non-excellent personnel are implemented, for example, steps S11 to S16 shown in FIG. 1 and steps S11 to S17 shown in FIG. 2.
  • the processor 30 executes the computer-readable instructions 40
  • the functions of the modules in the embodiment of the non-outstanding personnel training system described above are implemented, for example, modules 101 to 106 in FIG. 3.
  • the computer-readable instructions 40 may be divided into one or more modules / units, the one or more modules / units are stored in the memory 20 and executed by the processor 30, To complete this application.
  • the one or more modules / units may be a series of computer-readable instruction segments capable of performing specific functions, and the instruction segments are used to describe the execution process of the computer-readable instructions 40 in the computer device 1.
  • the computer-readable instructions 40 may be divided into an acquisition module 101, a establishment module 102, a first calculation module 103, a selection module 104, a second calculation module 105, and a determination module 106 in FIG. 3.
  • the second embodiment For specific functions of each module, refer to the second embodiment.
  • the computer device 1 may be a computing device such as a desktop computer, a notebook, a palmtop computer, and a cloud server.
  • a computing device such as a desktop computer, a notebook, a palmtop computer, and a cloud server.
  • the schematic diagram is only an example of the computer device 1, and does not constitute a limitation on the computer device 1. It may include more or fewer components than shown in the figure, or combine some components, or different Components, for example, the computer apparatus 1 may further include an input-output device, a network access device, a bus, and the like.
  • the so-called processor 30 may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application specific integrated circuits (ASICs), Ready-made programmable gate array (Field-Programmable Gate Array, FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc.
  • the general-purpose processor may be a microprocessor, or the processor 30 may be any conventional processor, etc.
  • the processor 30 is a control center of the computer device 1 and uses various interfaces and lines to connect the entire computer device 1 The various parts.
  • the memory 20 may be configured to store the computer-readable instructions 40 and / or modules / units, and the processor 30 may execute or execute the computer-readable instructions and / or modules / units stored in the memory 20, and
  • the data stored in the memory 20 is called to implement various functions of the computer device 1.
  • the memory 20 may mainly include a storage program area and a storage data area, where the storage program area may store an operating system, application programs required for at least one function (such as a sound playback function, an image playback function, etc.), etc .; the storage data area may Data (such as audio data, phone book, etc.) created according to the use of the computer device 1 are stored.
  • the memory 20 may include a high-speed random access memory, and may also include a non-volatile memory, such as a hard disk, an internal memory, a plug-in hard disk, a Smart Memory Card (SMC), and a Secure Digital (SD).
  • a non-volatile memory such as a hard disk, an internal memory, a plug-in hard disk, a Smart Memory Card (SMC), and a Secure Digital (SD).
  • SSD Secure Digital
  • flash memory card Flash card
  • flash memory device at least one disk storage device, flash memory device, or other volatile solid-state storage device.
  • the modules / units integrated in the computer device 1 When the modules / units integrated in the computer device 1 are implemented in the form of software functional units and sold or used as independent products, they can be stored in a non-volatile readable storage medium. Based on this understanding, this application implements all or part of the processes in the methods of the above embodiments, and can also be completed by computer-readable instructions instructing related hardware.
  • the computer-readable instructions can be stored in a non-volatile memory. In the read storage medium, when the computer-readable instructions are executed by a processor, the steps of the foregoing method embodiments can be implemented.
  • the computer-readable instructions include computer-readable instruction codes, and the computer-readable instruction codes may be in a source code form, an object code form, an executable file, or some intermediate form.
  • the non-volatile readable medium may include: any entity or device capable of carrying the computer-readable instruction code, a recording medium, a U disk, a mobile hard disk, a magnetic disk, an optical disk, a computer memory, a read-only memory (ROM, Read-Only Memory), Random Access Memory (RAM, Random Access Memory), electric carrier signals, telecommunication signals, and software distribution media.
  • ROM Read-Only Memory
  • RAM Random Access Memory
  • electric carrier signals telecommunication signals
  • telecommunication signals and software distribution media.
  • the content contained in the non-volatile readable medium may be appropriately increased or decreased according to the requirements of legislation and patent practice in the jurisdictions. For example, in some jurisdictions, according to legislation and patent practices, non- Volatile readable media does not include electrical carrier signals and telecommunication signals.

Landscapes

  • Engineering & Computer Science (AREA)
  • Business, Economics & Management (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Data Mining & Analysis (AREA)
  • Educational Technology (AREA)
  • Educational Administration (AREA)
  • General Physics & Mathematics (AREA)
  • Evolutionary Computation (AREA)
  • General Health & Medical Sciences (AREA)
  • General Engineering & Computer Science (AREA)
  • Tourism & Hospitality (AREA)
  • Strategic Management (AREA)
  • Health & Medical Sciences (AREA)
  • Human Resources & Organizations (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Artificial Intelligence (AREA)
  • Biophysics (AREA)
  • Evolutionary Biology (AREA)
  • Computing Systems (AREA)
  • Mathematical Physics (AREA)
  • Software Systems (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Computational Linguistics (AREA)
  • Molecular Biology (AREA)
  • Biomedical Technology (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Economics (AREA)
  • Marketing (AREA)
  • Primary Health Care (AREA)
  • General Business, Economics & Management (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

一种非绩优人员培训方法、系统、计算机装置及存储介质。所述非绩优人员培训方法包括：获取多个绩优人员的样本数据（S11）；根据多个绩优人员的非行为因子数据建立并训练得到一潜力绩优识别模型（S12）；将非绩优人员的非行为因子数据输入至所述潜力绩优识别模型计算得出所述非绩优人员成长为每一绩优类型的概率值（S13）；从计算得出的所有概率值中选取一最大概率值，并将最大概率值对应的绩优类型作为所述非绩优人员的潜力绩优类型（S14）；计算所述潜力绩优类型中的绩优人员的每一行为因子的均值与标准差，并将所述非绩优人员的每一行为因子值与所述潜力绩优类型的每一行为因子的均值进行一一对应比较（S15），当所述非绩优人员的一行为因子值与对应的行为因子的均值的差值大于标准差的预设倍数时，判定所述非绩优人员需要选择提升该行为因子的培训课程（S16）。本方法基于神经网络训练得到潜力绩优识别模型，根据模型定位出非绩优人员成长的潜力绩优类型，再对成长为该潜力绩优类型的行为因素进行有针对性培训。

Description

非绩优人员培训方法、系统、计算机装置及存储介质
本申请要求于2018年08月27日提交中国专利局，申请号为201810983357.4发明名称为“非绩优人员培训方法、系统、终端及计算机可读存储介质”的中国专利申请的优先权，其全部内容通过引用结合在本申请中。
技术领域
本申请涉及数据处理领域，尤其涉及一种非绩优人员培训方法、系统、计算机装置及存储介质。
背景技术
本部分旨在为权利要求书及具体实施方式中陈述的本申请的实施方式提供背景或上下文。此处的描述不因为包括在本部分中就承认是现有技术。
当今时代企业的发展主要表现在人力资源的竞争上，企业对员工的必要培训是人力资源开发的基础。现有人员培训方案主要可分为三类：一是统一进行人员培训，二是个人选择培训课程，三是他人推荐培训课程。第一类方法不能因人而异，对不同适应性水平的学习者采用相同的培训策略，难以因材施教。后两种方法太过主观，培训方案是否合适与选择人的判断是否正确相关很大。
发明内容
鉴于上述，本申请提供一种非绩优人员培训方法、系统、计算机装置及存储介质，可以实现根据不同的培训人员进行有针对性的培训方案生成。
本申请一实施方式提供一种非绩优人员培训方法，所述方法包括：获取多个绩优人员的样本数据，其中所述样本数据包括行为因子数据及非行为因子数据，多个所述绩优人员分属于多种绩优类型，每一所述绩优人员分属于一种绩优类型；
根据多个所述绩优人员的非行为因子数据建立并训练得到一潜力绩优识别模型；
将非绩优人员的非行为因子数据输入至所述潜力绩优识别模型计算得出所述非绩优人员成长为每一绩优类型的概率值；
从计算得出的所有概率值中选取一最大概率值，并将最大概率值对应的绩优类型作为所述非绩优人员的潜力绩优类型；
计算所述潜力绩优类型中的绩优人员的每一行为因子的均值与标准差，并将所述非绩优人员的每一行为因子值与所述潜力绩优类型的每一行为因子的均值进行一一对应比较；及
当所述非绩优人员的一行为因子值与对应的行为因子的均值的差值大于标准差的预设倍数时，判定所述非绩优人员需要选择提升该行为因子的培训 课程。
本申请一实施方式提供一种非绩优人员培训系统，所述系统包括：
获取模块，用于获取多个绩优人员的样本数据，其中所述样本数据包括行为因子数据及非行为因子数据，多个所述绩优人员分属于多种绩优类型，每一所述绩优人员分属于一种绩优类型；
建立模块，用于根据多个所述绩优人员的非行为因子数据建立并训练得到一潜力绩优识别模型；
第一计算模块，用于将非绩优人员的非行为因子数据输入至所述潜力绩优识别模型计算得出所述非绩优人员成长为每一绩优类型的概率值；
选取模块，用于从计算得出的所有概率值中选取一最大概率值，并将最大概率值对应的绩优类型作为所述非绩优人员的潜力绩优类型；
第二计算模块，用于计算所述潜力绩优类型中的绩优人员的每一行为因子的均值与标准差，并将所述非绩优人员的每一行为因子值与所述潜力绩优类型的每一行为因子的均值进行一一对应比较；及
判定模块，用于在所述非绩优人员的一行为因子值与对应的行为因子的均值的差值大于标准差的预设倍数时，判定所述非绩优人员需要选择提升该行为因子的培训课程。
本申请一实施方式提供一种计算机装置，所述计算机装置包括处理器及存储器，所述存储器上存储有若干计算机可读指令，所述处理器用于执行存储器中存储的计算机可读指令时实现如前面所述的非绩优人员培训方法的步骤。
本申请一实施方式提供一种非易失性可读存储介质，其上存储有计算机可读指令，所述计算机可读指令被处理器执行时实现如前面所述的非绩优人员培训方法的步骤。
上述非绩优人员培训方法、系统、计算机装置及非易失性可读存储介质，基于机器学习与绩优人员的非行为因子数据来建立并训练得到潜力绩优识别模型，并利用该模型实现识别出非绩优人员成长的潜力的绩优类型，将非绩优人员的一行为因子与绩优人员的行为因子均值进行比较来判定非绩优人员是否需要选择提升该行为因子的培训课程，实现从先天的角度客观确定非绩优人员最有可能成长的绩优类型，再从后天的角度找到了非绩优人员与目标绩优类型的不足，提升人员培训效果。
附图说明
图1是本申请一实施例中非绩优人员培训方法的步骤流程图。
图2是本申请另一实施例中非绩优人员培训方法的步骤流程图。
图3为本申请一实施例中非绩优人员培训系统的功能模块图。
图4为本申请一实施例中计算机装置示意图。
具体实施方式
优选地，本申请的非绩优人员培训方法应用在一个或者多个计算机装置 中。所述计算机装置是一种能够按照事先设定或存储的指令，自动进行数值计算和/或信息处理的设备，其硬件包括但不限于微处理器、专用集成电路(Application Specific Integrated Circuit，ASIC)、可编程门阵列(Field－Programmable Gate Array，FPGA)、数字处理器(Digital Signal Processor，DSP)、嵌入式设备等。
实施例一：
图1是本申请非绩优人员培训方法较佳实施例的步骤流程图。根据不同的需求，所述流程图中步骤的顺序可以改变，某些步骤可以省略。
参阅图1所示，所述非绩优人员培训方法具体包括以下步骤。
步骤S11、获取多个绩优人员的样本数据，其中所述样本数据包括行为因子数据及非行为因子数据，多个所述绩优人员分属于多种绩优类型，每一所述绩优人员分属于一种绩优类型。
在一实施方式中，可以通过接入网络来连接至一绩优人员样本库，进而来获取所述绩优人员样本库存储的绩优人员的样本数据。所述绩优人员样本库可以通过大数据方式搜集多个绩优人员的样本数据，也可以通过接收人为录入的多个绩优人员的样本数据。
在一实施方式中，多个所述绩优人员分属于多种绩优类型，但每一所述绩优人员只分属于一个绩优类型，即每一所述绩优人员不能同时属于多种绩优类型。
举例而言，该多个绩优类型可以包括三种绩优类型，分别为资源型、学习型及勤奋型。资源型可以是指业务能力强、工作能力强的人员，学习型可以是指学习能力强、成长能力强的人员，勤奋型可以是指学习时间长、每天工作时间长的人员。对于每一绩优人员来说，其只能被归类到三种绩优类型中的一种。以下以绩优人员和非绩优人员是A公司员工为例进行举例说明。
举例而言，A公司包括1000名员工，该1000名员工可以被划分为绩优人员和非绩优人员两种类型，每一员工只能被划分为绩优人员或者被划分为非绩优人员，绩优人员和非绩优人员的划分可以根据多个绩优类型对应的划分规则进行划分，比如对于学习型绩优类型，可以根据每一员工的学习能力、成长能力来判定某一员工是否为学习型的绩优人员，再比如对应资源型绩优类型，可以根据每一员工的业务能力来判定某一员工是否属于资源型的绩优人员。绩优人员的类型包括资源型、学习型及勤奋型。比如根据每一员工的日常表现，其中200名员工被划分为绩优人员，其余800名员工被划分为非绩优人员。在200名绩优人员中，每一绩优人员属于一种绩优类型，资源型绩优人员包括80人，学习型绩优人员包括50人，勤奋型绩优人员包括70人。
在一实施方式中，绩优人员和非绩优人员的样本数据均包括行为因子数据及非行为因子数据。所述行为因子数据包括多个行为因子，所述非行为因子数据包括多个非行为因子，每一所述行为因子对应有一行为因子值，每一所述非行为因子对应有一非行为因子值，且所述绩优人员与所述非绩优人员 具有相同的行为因子与非行为因子。所述行为因子优选是指后天的行为因素或者短时间可以改变、培养的行为因素，比如所述行为因子包括活动范围、出入各场所的频次、考勤、APP活跃情况、沟通能力等。所述非行为因子优选是指先天的行为因素或者短时间难以改变、培养的行为因素，比如所述非行为因子包括年龄、性别、学历等。
举例而言，一非行为因子为年龄，则对应的非行为因子值为年龄大小，若绩优人员A1的年龄为27，绩优人员A2的年龄为30，则绩优人员A1的年龄的非行为因子值为27，绩优人员A2的年龄的非行为因子值为30；若一行为因子为沟通能力，该行为因子的行为因子值可以根据员工的实际沟通能力水平进行评分(可以根据一预设评分标准)得到对应的行为因子值。在步骤S11中，所述获取多个绩优人员的样本数据即获取多个绩优人员的行为因子数据及非行为因子数据。
步骤S12、根据多个所述绩优人员的非行为因子数据建立并训练得到一潜力绩优识别模型。
在一实施方式中，所述潜力绩优识别模型可以是基于神经网络模型和多个所述绩优人员的非行为因子数据训练出来的分类模型。具体地，可以先建立一神经网络模型，所述神经网络模型包括输入层、多个隐藏层及输出层，再利用多个所述绩优人员的非行为因子数据对所述神经网络模型进行训练得到所述潜力绩优识别模型。
所述神经网络模型的输入层用于接收多个所述绩优人员的非行为因子数据，每一隐藏层包括多个节点(神经元)，每一隐藏层中的每一节点被配置成对来自该模型中的相邻下层的至少一个节点的输出执行线性或非线性变换。其中，上层隐藏层的节点的输入可以基于相邻下层中的一个节点或若干节点的输出，每个隐藏层具有对应的权值，该权值是基于训练样本数据获得的。在对该模型进行训练时，可以通过利用有监督的学习过程来进行模型的训练，得到各个隐藏层的初始权值。可以通过向后传播(Back propagation，BP)算法来对各隐藏层的权值的进行调节，所述神经网络模型的输出层用于接收来自最后一层隐藏层的输出信号。
在一实施方式中，所述步骤S12可以具体包括：
a.将多个所述绩优人员的非行为因子数据划分为训练集及验证集；
b.建立一神经网络模型，并利用所述训练集对所述神经网络模型进行训练；
c.利用所述验证集对训练后的神经网络模型进行验证，并根据每一验证结果统计得到一模型预测准确率；
d.判断所述模型预测准确率是否小于预设阈值；
e.若所述模型预测准确率不小于所述预设阈值，将训练完成的所述神经网络模型作为所述潜力绩优识别模型。
f.若所述模型预测准确率小于所述预设阈值，调整所述神经网络模型的参数，并利用所述训练集重新对调整后的神经网络模型进行训练，直到验证 集验证得到的模型预测准确率不小于所述预设阈值，其中所述神经网络模型的参数包括总层数、每一层的神经元数等。
在一实施方式中，所述调整所述神经网络模型的参数可以是调整所述神经网络模型的总层数和/或每一层的神经元数。所述训练集用于对神经网络模型进行训练，所述验证集用于对训练后的神经网络模型进行验证。
在一实施方式中，可以先利用所述训练集对神经网络模型进行训练得到一中间模型，再将所述验证集中的绩优人员非行为因子数据输入至所述中间模型中进行绩优类型分类验证，并根据每一验证结果可以统计得到一模型预测准确率；再判断所述中间模型预测准确率是否小于预设阈值；若所述中间模型预测准确率不小于所述预设阈值，表明此中间模型分类效果较好，满足使用需求，可以直接将所述中间模型作为所述潜力绩优识别模型；若所述模型预测准确率小于所述预设阈值时，表明此中间模型分类效果不好，需要进行改善，此时可以调整所述神经网络模型的参数，并利用所述训练集重新对调整后的神经网络模型进行训练得到一新的中间模型，然后再次利用所述验证集对重新得到的中间模型进行验证得到一新的模型预测准确率，再判断该新的模型预测准确率是否小于预设阈值，若该新的模型预测准确率不小于所述预设阈值，表明重新得到的中间模型分类效果较好，满足使用需求，可以将重新得到的中间模型作为所述潜力绩优识别模型；如果该新的模型预测准确率仍然小于所述预设阈值，需要再次重复上述步骤直至通过验证集得到的模型预测准确率不小于所述预设阈值。
在一实施方式中，所述预设阈值可以根据实际使用需求进行设定，例如所述预设阈值设置为95％，即模型预测准确率需不小于95％。
步骤S13、将非绩优人员的非行为因子数据输入至所述潜力绩优识别模型计算得出所述非绩优人员成长为每一绩优类型的概率值。
在一实施方式中，通过步骤S12的训练与验证，所述潜力绩优识别模型可以实现对非绩优人员的潜力绩优类型进行识别，此时可以将非绩优人员的非行为因子数据作为所述潜力绩优识别模型的输入，将所述潜力绩优识别模型的输出视为非绩优人员成长为各绩优类型的概率。
S14、从计算得出的所有概率值中选取一最大概率值，并将最大概率值对应的绩优类型作为所述非绩优人员的潜力绩优类型。
在一实施方式中，每一所述概率值代表着所述非绩优人员成长为每一绩优类型的绩优人员的概率值，可以从计算得出的所有概率值中选取一最大概率值，并将最大概率值对应的绩优类型作为所述非绩优人员的潜力绩优类型，从而实现识别出所述非绩优人员成长的最具潜力的绩优类型。
举例而言，通过所述潜力绩优识别模型得到非绩优人员A1成长为资源型绩优人员的概率为0.6，成长为学习型绩优人员的概率为0.7，成长为勤奋型绩优人员的概率为0.55，由于成长为学习型绩优人员的概率值最大，则将学习型判定为该非绩优人员A1的潜力绩优类型。
在一实施方式中，当存在多个最大概率值时，可以随机选择一最大概率 值对应的绩优类型作为所述非绩优人员的潜力绩优类型。比如，通过所述潜力绩优识别模型得到非绩优人员A2成长为资源型绩优人员的概率为0.8，成长为学习型绩优人员的概率为0.7，成长为勤奋型绩优人员的概率为0.8，由于成长为勤奋型绩优人员与成长为资源型绩优人员的概率值均为0.8，则可以选择将勤奋型判定为该非绩优人员A2的潜力发展绩优类型，也可以将资源型判定为该非绩优人员A2的潜力发展绩优类型。
在一实施方式中，当存在多个最大概率值时，还可以选择一与预设需求匹配的最大概率值对应的绩优类型作为所述非绩优人员的潜力绩优类型。比如，所述预设需求可以是公司当前紧要人员类型需求，若公司对于资源型绩优人员需求最大，对学习型绩优人员需求次之，对勤奋型绩优人员需求最低。此时若通过所述潜力绩优识别模型得到非绩优人员A2成长为资源型绩优人员的概率为0.8，成长为学习型绩优人员的概率为0.7，成长为勤奋型绩优人员的概率为0.8，由于成长为勤奋型绩优人员与成长为资源型绩优人员的概率值均为0.8且公司对于资源型绩优人员需求最大，则可以将资源型判定为该非绩优人员A2的潜力发展绩优类型。
S15、计算所述潜力绩优类型中的绩优人员的每一行为因子的均值与标准差，并将所述非绩优人员的每一行为因子值与所述潜力绩优类型的每一行为因子的均值进行一一对应比较。
在一实施方式中，通过步骤S14可以确定非绩优人员的潜力绩优类型，该潜力绩优类型包含有多个绩优人员的行为因子数据，通过计算所述潜力绩优类型中的绩优人员的每一行为因子的均值与标准差，再将所述非绩优人员的每一行为因子值与所述潜力绩优类型的每一行为因子的均值进行一一对应比较来实现判断所述非绩优人员的每一行为因子是否需要进行提升。
在一实施方式中，所述非绩优人员的每一行为因子预先已经过量化，每一行为因子均具有一行为因子值。同样地，每一绩优人员的每一行为因子预先也已经过量化，即每一绩优人员的每一行为因子均具有一行为因子值。假设每一人员(绩优和非绩优)具有50个行为因子，通过步骤S14得到非绩优人员A1的潜力绩优类型是资源型。在绩优人员样本库中，资源型有100个绩优人员，则对于每一行为因子均具有100个值，求这100个绩优人员的每一行为因子的均值和标准差。比如，对于100个绩优人员的行为因子a1、行为因子a2而言，计算得到行为因子a1在该100个绩优人员的行为因子均值为70，标准差为4，计算得到行为因子a2在该100个绩优人员的行为因子均值为72，标准差为3.5。再将所述非绩优人员的行为因子a1的值与该均值70进行比较，将所述非绩优人员的行为因子a2的值与该均值72进行比较。
步骤S16、当所述非绩优人员的一行为因子值与对应的行为因子的均值的差值大于标准差的预设倍数时，判定所述非绩优人员需要选择提升该行为因子的培训课程。
在一实施方式中，所述预设倍数可以根据实际使用场景进行设定，比如设定为2倍的标准差。在本申请的其他实施方式中，也可以设定为3倍或者 4倍的标准差。
在一实施方式中，以2倍的标准差进行举例说明，所述非绩优人员的行为因子a1的值与对应均值70进行比较，判断差值是否大于2倍的标准差，即判断所述非绩优人员的行为因子a1的值与对应均值70的差值是否大于8(2*4)，若差值大于8，则判断所述非绩优人员需要选择提升该行为因子a1的培训课程。将所述非绩优人员的行为因子a2的值与对应均值72进行比较，判断差值是否大于2倍的标准差，即判断所述非绩优人员的行为因子a2的值与对应均值72的差值是否大于7(2*3.5)，若差值大于7，则判断所述非绩优人员需要选择提升该行为因子a2的培训课程。
请同时参阅图2，与图1示出的非绩优人员培训方法相比，图2示处的非绩优人员培训方法还包括步骤S17。
步骤S17，当所述非绩优人员的一行为因子值与对应的行为因子的均值的差值不大于标准差的预设倍数时，判定所述非绩优人员不需要选择提升该行为因子的培训课程。
在一实施方式中，同样以2倍的标准差进行举例说明，所述非绩优人员的行为因子a1的值与对应均值70进行比较，判断差值是否大于2倍的标准差，即判断所述非绩优人员的行为因子a1的值与对应均值70的差值是否大于8(2*4)，若差值不大于8，则判断所述非绩优人员不需要选择提升该行为因子a1的培训课程。将所述非绩优人员的行为因子a2的值与对应均值72进行比较，判断差值是否大于2倍的标准差，即判断所述非绩优人员的行为因子a2的值与对应均值72的差值是否大于7(2*3.5)，若差值不大于7，则判断所述非绩优人员不需要选择提升该行为因子a2的培训课程。
举例而言，一非绩优人员A1的行为因子a1的值为63，行为因子a2的值为60。此时可计算得到所述非绩优人员A1的行为因子a1与对应均值的差值＝(70-63)＝7，其小于2倍的标准差(2*4)，则判断所述非绩优人员A1暂时不需要选择提升该行为因子a1的培训课程。所述非绩优人员A1的行为因子a2与对应均值的差值＝(72-60)＝12，其大于2倍的标准差(2*3.5)，则判断所述非绩优人员A1需要选择提升该行为因子a2的培训课程。
实施例二：
图3为本申请非绩优人员培训系统较佳实施例的功能模块图。
参阅图2所示，所述非绩优人员培训系统10可以包括获取模块101、建立模块102、第一计算模块103、选取模块104、第二计算模块105及判定模块106。
获取模块101用于获取多个绩优人员的样本数据，其中所述样本数据包括行为因子数据及非行为因子数据，多个所述绩优人员分属于多种绩优类型，每一所述绩优人员分属于一种绩优类型。
在一实施方式中，所述获取模块101可以通过接入网络来连接至一绩优人员样本库，进而来获取所述绩优人员样本库存储的绩优人员的样本数据。所述绩优人员样本库可以通过大数据方式搜集多个绩优人员的样本数据，也 可以通过接收人为录入的多个绩优人员的样本数据。
在一实施方式中，多个所述绩优人员分属于多种绩优类型，但每一所述绩优人员只分属于一个绩优类型，即每一所述绩优人员不能同时属于多种绩优类型。
举例而言，该多个绩优类型可以包括三种绩优类型，分别为资源型、学习型及勤奋型。资源型可以是指业务能力强、工作能力强的人员，学习型可以是指学习能力强、成长能力强的人员，勤奋型可以是指学习时间长、每天工作时间长的人员。对于每一绩优人员来说，其只能被归类到三种绩优类型中的一种。以下以绩优人员和非绩优人员是A公司员工为例进行举例说明。
举例而言，A公司包括1000名员工，该1000名员工可以被划分为绩优人员和非绩优人员两种类型，每一员工只能被划分为绩优人员或者被划分为非绩优人员，绩优人员和非绩优人员的划分可以根据多个绩优类型对应的划分规则进行划分，比如对于学习型绩优类型，可以根据每一员工的学习能力、成长能力来判定某一员工是否为学习型的绩优人员，再比如对应资源型绩优类型，可以根据每一员工的业务能力来判定某一员工是否属于资源型的绩优人员。绩优人员的类型包括资源型、学习型及勤奋型。比如根据每一员工的日常表现，其中200名员工被划分为绩优人员，其余800名员工被划分为非绩优人员。在200名绩优人员中，每一绩优人员属于一种绩优类型，资源型绩优人员包括80人，学习型绩优人员包括50人，勤奋型绩优人员包括70人。
在一实施方式中，绩优人员和非绩优人员的样本数据均包括行为因子数据及非行为因子数据。所述行为因子数据包括多个行为因子，所述非行为因子数据包括多个非行为因子，每一所述行为因子对应有一行为因子值，每一所述非行为因子对应有一非行为因子值，且所述绩优人员与所述非绩优人员具有相同的行为因子与非行为因子。所述行为因子优选是指后天的行为因素或者短时间可以改变、培养的行为因素，比如所述行为因子包括活动范围、出入各场所的频次、考勤、APP活跃情况、沟通能力等。所述非行为因子优选是指先天的行为因素或者短时间难以改变、培养的行为因素，比如所述非行为因子包括年龄、性别、学历等。
举例而言，一非行为因子为年龄，则对应的非行为因子值为年龄大小，若绩优人员A1的年龄为27，绩优人员A2的年龄为30，则绩优人员A1的年龄的非行为因子值为27，绩优人员A2的年龄的非行为因子值为30；若一行为因子为沟通能力，该行为因子的行为因子值可以根据员工的实际沟通能力水平进行评分(可以根据一预设评分标准)得到对应的行为因子值。在步骤S11中，所述获取多个绩优人员的样本数据即获取多个绩优人员的行为因子数据及非行为因子数据。
建立模块102用于根据多个所述绩优人员的非行为因子数据建立并训练得到一潜力绩优识别模型。
在一实施方式中，所述潜力绩优识别模型可以是基于神经网络模型和多 个所述绩优人员的非行为因子数据训练出来的分类模型。具体地，所述建立模块102可以先建立一神经网络模型，所述神经网络模型包括输入层、多个隐藏层及输出层，再利用多个所述绩优人员的非行为因子数据对所述神经网络模型进行训练得到所述潜力绩优识别模型。
所述神经网络模型的输入层用于接收多个所述绩优人员的非行为因子数据，每一隐藏层包括多个节点(神经元)，每一隐藏层中的每一节点被配置成对来自该模型中的相邻下层的至少一个节点的输出执行线性或非线性变换。其中，上层隐藏层的节点的输入可以基于相邻下层中的一个节点或若干节点的输出，每个隐藏层具有对应的权值，该权值是基于训练样本数据获得的。在对该模型进行训练时，可以通过利用有监督的学习过程来进行模型的训练，得到各个隐藏层的初始权值。可以通过向后传播(Back propagation，BP)算法来对各隐藏层的权值的进行调节，所述神经网络模型的输出层用于接收来自最后一层隐藏层的输出信号。
在一实施方式中，所述建立模块102建立并训练得到一潜力绩优识别模型的方式可以具体包括：
a.将多个所述绩优人员的非行为因子数据划分为训练集及验证集；
b.建立一神经网络模型，并利用所述训练集对所述神经网络模型进行训练；
c.利用所述验证集对训练后的神经网络模型进行验证，并根据每一验证结果统计得到一模型预测准确率；
d.判断所述模型预测准确率是否小于预设阈值；
e.若所述模型预测准确率不小于所述预设阈值，将训练完成的所述神经网络模型作为所述潜力绩优识别模型。
f.若所述模型预测准确率小于所述预设阈值，调整所述神经网络模型的参数，并利用所述训练集重新对调整后的神经网络模型进行训练，直到验证集验证得到的模型预测准确率不小于所述预设阈值，其中所述神经网络模型的参数包括总层数、每一层的神经元数等。
在一实施方式中，所述调整所述神经网络模型的参数可以是调整所述神经网络模型的总层数和/或每一层的神经元数。所述训练集用于对神经网络模型进行训练，所述验证集用于对训练后的神经网络模型进行验证。
在一实施方式中，所述建立模块102可以先利用所述训练集对神经网络模型进行训练得到一中间模型，再将所述验证集中的绩优人员非行为因子数据输入至所述中间模型中进行绩优类型分类验证，并根据每一验证结果可以统计得到一模型预测准确率；再判断所述中间模型预测准确率是否小于预设阈值；若所述中间模型预测准确率不小于所述预设阈值，表明此中间模型分类效果较好，满足使用需求，可以直接将所述中间模型作为所述潜力绩优识别模型；若所述模型预测准确率小于所述预设阈值时，表明此中间模型分类效果不好，需要进行改善，此时可以调整所述神经网络模型的参数，并利用所述训练集重新对调整后的神经网络模型进行训练得到一新的中间模型，然 后再次利用所述验证集对重新得到的中间模型进行验证得到一新的模型预测准确率，再判断该新的模型预测准确率是否小于预设阈值，若该新的模型预测准确率不小于所述预设阈值，表明重新得到的中间模型分类效果较好，满足使用需求，可以将重新得到的中间模型作为所述潜力绩优识别模型；如果该新的模型预测准确率仍然小于所述预设阈值，需要再次重复上述步骤直至通过验证集得到的模型预测准确率不小于所述预设阈值。
在一实施方式中，所述预设阈值可以根据实际使用需求进行设定，例如所述预设阈值设置为95％，即模型预测准确率需不小于95％。
所述第一计算模块103用于将非绩优人员的非行为因子数据输入至所述潜力绩优识别模型计算得出所述非绩优人员成长为每一绩优类型的概率值。
在一实施方式中，通过所述建立模块102的训练与验证，所述潜力绩优识别模型可以实现对非绩优人员的潜力绩优类型进行识别，此时可以将非绩优人员的非行为因子数据作为所述潜力绩优识别模型的输入，将所述潜力绩优识别模型的输出视为非绩优人员成长为各绩优类型的概率。
选取模块104用于从计算得出的所有概率值中选取一最大概率值，并将最大概率值对应的绩优类型作为所述非绩优人员的潜力绩优类型。
在一实施方式中，每一所述概率值代表着所述非绩优人员成长为每一绩优类型的绩优人员的概率值，所述选取模块104可以从计算得出的所有概率值中选取一最大概率值，并将最大概率值对应的绩优类型作为所述非绩优人员的潜力绩优类型，从而实现识别出所述非绩优人员成长的最具潜力的绩优类型。
举例而言，通过所述潜力绩优识别模型得到非绩优人员A1成长为资源型绩优人员的概率为0.6，成长为学习型绩优人员的概率为0.7，成长为勤奋型绩优人员的概率为0.55，由于成长为学习型绩优人员的概率值最大，则将学习型判定为该非绩优人员A1的潜力绩优类型。
在一实施方式中，当存在多个最大概率值时，所述选取模块104可以随机选择一最大概率值对应的绩优类型作为所述非绩优人员的潜力绩优类型。比如，通过所述潜力绩优识别模型得到非绩优人员A2成长为资源型绩优人员的概率为0.8，成长为学习型绩优人员的概率为0.7，成长为勤奋型绩优人员的概率为0.8，由于成长为勤奋型绩优人员与成长为资源型绩优人员的概率值均为0.8，则可以选择将勤奋型判定为该非绩优人员A2的潜力发展绩优类型，也可以将资源型判定为该非绩优人员A2的潜力发展绩优类型。
在一实施方式中，当存在多个最大概率值时，所述选取模块104还可以选择一与预设需求匹配的最大概率值对应的绩优类型作为所述非绩优人员的潜力绩优类型。比如，所述预设需求可以是公司当前紧要人员类型需求，若公司对于资源型绩优人员需求最大，对学习型绩优人员需求次之，对勤奋型绩优人员需求最低。此时若通过所述潜力绩优识别模型得到非绩优人员A2成长为资源型绩优人员的概率为0.8，成长为学习型绩优人员的概率为0.7，成长为勤奋型绩优人员的概率为0.8，由于成长为勤奋型绩优人员与成长为 资源型绩优人员的概率值均为0.8且公司对于资源型绩优人员需求最大，则可以将资源型判定为该非绩优人员A2的潜力发展绩优类型。
第二计算模块105用于计算所述潜力绩优类型中的绩优人员的每一行为因子的均值与标准差，并将所述非绩优人员的每一行为因子值与所述潜力绩优类型的每一行为因子的均值进行一一对应比较。
在一实施方式中，通过所述选取模块104可以确定非绩优人员的潜力绩优类型，该潜力绩优类型包含有多个绩优人员的行为因子数据，通过所述第二计算模块105可以计算得到所述潜力绩优类型中的绩优人员的每一行为因子的均值与标准差，再将所述非绩优人员的每一行为因子值与所述潜力绩优类型的每一行为因子的均值进行一一对应比较来实现判断所述非绩优人员的每一行为因子是否需要进行提升。
在一实施方式中，所述非绩优人员的每一行为因子预先已经过量化，每一行为因子均具有一行为因子值。同样地，每一绩优人员的每一行为因子预先也已经过量化，即每一绩优人员的每一行为因子均具有一行为因子值。假设每一人员(绩优和非绩优)具有50个行为因子，通过步骤S14得到非绩优人员A1的潜力绩优类型是资源型。在绩优人员样本库中，资源型有100个绩优人员，则对于每一行为因子均具有100个值，求这100个绩优人员的每一行为因子的均值和标准差。比如，对于100个绩优人员的行为因子a1、行为因子a2而言，计算得到行为因子a1在该100个绩优人员的行为因子均值为70，标准差为4，计算得到行为因子a2在该100个绩优人员的行为因子均值为72，标准差为3.5。再将所述非绩优人员的行为因子a1的值与该均值70进行比较，将所述非绩优人员的行为因子a2的值与该均值72进行比较。
判定模块106用于在所述非绩优人员的一行为因子值与对应的行为因子的均值的差值大于标准差的预设倍数时，判定所述非绩优人员需要选择提升该行为因子的培训课程。
在一实施方式中，所述预设倍数可以根据实际使用场景进行设定，比如设定为2倍的标准差。在本申请的其他实施方式中，也可以设定为3倍或者4倍的标准差。
在一实施方式中，以2倍的标准差进行举例说明，所述非绩优人员的行为因子a1的值与对应均值70进行比较，判断差值是否大于2倍的标准差，即判断所述非绩优人员的行为因子a1的值与对应均值70的差值是否大于8(2*4)，若差值大于8，则所述判定模块106判断所述非绩优人员需要选择提升该行为因子a1的培训课程。将所述非绩优人员的行为因子a2的值与对应均值72进行比较，判断差值是否大于2倍的标准差，即判断所述非绩优人员的行为因子a2的值与对应均值72的差值是否大于7(2*3.5)，若差值大于7，则所述判定模块106判断所述非绩优人员需要选择提升该行为因子a2的培训课程。
在一实施方式中，所述判定模块106还用于在所述非绩优人员的一行为因子值与对应的行为因子的均值的差值不大于标准差的预设倍数时，判定所 述非绩优人员不需要选择提升该行为因子的培训课程。
在一实施方式中，同样以2倍的标准差进行举例说明，所述非绩优人员的行为因子a1的值与对应均值70进行比较，判断差值是否大于2倍的标准差，即判断所述非绩优人员的行为因子a1的值与对应均值70的差值是否大于8(2*4)，若差值不大于8，则所述判定模块106判断所述非绩优人员不需要选择提升该行为因子a1的培训课程。将所述非绩优人员的行为因子a2的值与对应均值72进行比较，判断差值是否大于2倍的标准差，即判断所述非绩优人员的行为因子a2的值与对应均值72的差值是否大于7(2*3.5)，若差值不大于7，则所述判定模块106判断所述非绩优人员不需要选择提升该行为因子a2的培训课程。
举例而言，一非绩优人员A1的行为因子a1的值为63，行为因子a2的值为60。此时可计算得到所述非绩优人员A1的行为因子a1与对应均值的差值＝(70-63)＝7，其小于2倍的标准差(2*4)，则所述判定模块106判断所述非绩优人员A1暂时不需要选择提升该行为因子a1的培训课程。所述非绩优人员A1的行为因子a2与对应均值的差值＝(72-60)＝12，其大于2倍的标准差(2*3.5)，则所述判定模块106判断所述非绩优人员A1需要选择提升该行为因子a2的培训课程。
图4为本申请计算机装置较佳实施例的示意图。
所述计算机装置1包括存储器20、处理器30以及存储在所述存储器20中并可在所述处理器30上运行的计算机可读指令40，例如非绩优人员培训程序。所述处理器30执行所述计算机可读指令40时实现上述非绩优人员培训方法实施例中的步骤，例如图1所示的步骤S11～S16、图2所示的步骤S11～S17。或者，所述处理器30执行所述计算机可读指令40时实现上述非绩优人员培训系统实施例中各模块的功能，例如图3中的模块101～106。
示例性的，所述计算机可读指令40可以被分割成一个或多个模块/单元，所述一个或者多个模块/单元被存储在所述存储器20中，并由所述处理器30执行，以完成本申请。所述一个或多个模块/单元可以是能够完成特定功能的一系列计算机可读指令段，所述指令段用于描述所述计算机可读指令40在所述计算机装置1中的执行过程。例如，所述计算机可读指令40可以被分割成图3中的获取模块101、建立模块102、第一计算模块103、选取模块104、第二计算模块105及判定模块106。各模块具体功能参见实施例二。
所述计算机装置1可以是桌上型计算机、笔记本、掌上电脑及云端服务器等计算设备。本领域技术人员可以理解，所述示意图仅仅是计算机装置1的示例，并不构成对计算机装置1的限定，可以包括比图示更多或更少的部件，或者组合某些部件，或者不同的部件，例如所述计算机装置1还可以包括输入输出设备、网络接入设备、总线等。
所称处理器30可以是中央处理单元(Central Processing Unit，CPU)，还可以是其他通用处理器、数字信号处理器(Digital Signal Processor，DSP)、专用集成电路(Application Specific Integrated Circuit，ASIC)、现成可编程门 阵列(Field-Programmable Gate Array，FPGA)或者其他可编程逻辑器件、分立门或者晶体管逻辑器件、分立硬件组件等。通用处理器可以是微处理器或者所述处理器30也可以是任何常规的处理器等，所述处理器30是所述计算机装置1的控制中心，利用各种接口和线路连接整个计算机装置1的各个部分。
所述存储器20可用于存储所述计算机可读指令40和/或模块/单元，所述处理器30通过运行或执行存储在所述存储器20内的计算机可读指令和/或模块/单元，以及调用存储在存储器20内的数据，实现所述计算机装置1的各种功能。所述存储器20可主要包括存储程序区和存储数据区，其中，存储程序区可存储操作系统、至少一个功能所需的应用程序(比如声音播放功能、图像播放功能等)等；存储数据区可存储根据计算机装置1的使用所创建的数据(比如音频数据、电话本等)等。此外，存储器20可以包括高速随机存取存储器，还可以包括非易失性存储器，例如硬盘、内存、插接式硬盘，智能存储卡(Smart Media Card,SMC)，安全数字(Secure Digital,SD)卡，闪存卡(Flash Card)、至少一个磁盘存储器件、闪存器件、或其他易失性固态存储器件。
所述计算机装置1集成的模块/单元如果以软件功能单元的形式实现并作为独立的产品销售或使用时，可以存储在一个非易失性可读取存储介质中。基于这样的理解，本申请实现上述实施例方法中的全部或部分流程，也可以通过计算机可读指令来指令相关的硬件来完成，所述的计算机可读指令可存储于一非易失性可读存储介质中，所述计算机可读指令在被处理器执行时，可实现上述各个方法实施例的步骤。其中，所述计算机可读指令包括计算机可读指令代码，所述计算机可读指令代码可以为源代码形式、对象代码形式、可执行文件或某些中间形式等。所述非易失性可读介质可以包括：能够携带所述计算机可读指令代码的任何实体或装置、记录介质、U盘、移动硬盘、磁碟、光盘、计算机存储器、只读存储器(ROM，Read-Only Memory)、随机存取存储器(RAM，Random Access Memory)、电载波信号、电信信号以及软件分发介质等。需要说明的是，所述非易失性可读介质包含的内容可以根据司法管辖区内立法和专利实践的要求进行适当的增减，例如在某些司法管辖区，根据立法和专利实践，非易失性可读介质不包括电载波信号和电信信号。
最后应说明的是，以上实施例仅用以说明本申请的技术方案而非限制，尽管参照较佳实施例对本申请进行了详细说明，本领域的普通技术人员应当理解，可以对本申请的技术方案进行修改或等同替换，而不脱离本申请技术方案的精神和范围。

Claims (20)

  1. 一种非绩优人员培训方法，其特征在于，所述方法包括：
    获取多个绩优人员的样本数据，其中所述样本数据包括行为因子数据及非行为因子数据，多个所述绩优人员分属于多种绩优类型，每一所述绩优人员分属于一种绩优类型；
    根据多个所述绩优人员的非行为因子数据建立并训练得到一潜力绩优识别模型；
    将非绩优人员的非行为因子数据输入至所述潜力绩优识别模型计算得出所述非绩优人员成长为每一绩优类型的概率值；
    从计算得出的所有概率值中选取一最大概率值，并将最大概率值对应的绩优类型作为所述非绩优人员的潜力绩优类型；
    计算所述潜力绩优类型中的绩优人员的每一行为因子的均值与标准差，并将所述非绩优人员的每一行为因子值与所述潜力绩优类型的每一行为因子的均值进行一一对应比较；及
    当所述非绩优人员的一行为因子值与对应的行为因子的均值的差值大于标准差的预设倍数时，判定所述非绩优人员需要选择提升该行为因子的培训课程。
  2. 如权利要求1所述的非绩优人员培训方法，其特征在于，所述行为因子数据包括多个行为因子，所述非行为因子数据包括多个非行为因子，每一所述行为因子对应有一行为因子值，每一所述非行为因子对应有一非行为因子值，且所述绩优人员与所述非绩优人员具有相同的行为因子与非行为因子。
  3. 如权利要求1或2所述的非绩优人员培训方法，其特征在于，所述根据多个所述绩优人员的非行为因子数据建立并训练得到一潜力绩优识别模型的步骤包括：
    建立一神经网络模型，所述神经网络模型包括输入层、多个隐藏层及输出层；及
    利用多个所述绩优人员的非行为因子数据对所述神经网络模型进行训练得到所述潜力绩优识别模型。
  4. 如权利要求1或2所述的非绩优人员培训方法，其特征在于，所述根据多个所述绩优人员的非行为因子数据建立并训练得到一潜力绩优识别模型的步骤包括：
    将多个所述绩优人员的非行为因子数据划分为训练集及验证集；
    建立一神经网络模型，并利用所述训练集对所述神经网络模型进行训练；
    利用所述验证集对训练后的神经网络模型进行验证，并根据每一验证结果统计得到一模型预测准确率；
    判断所述模型预测准确率是否小于预设阈值；
    若所述模型预测准确率不小于所述预设阈值，将训练完成的所述神经网 络模型作为所述潜力绩优识别模型。
  5. 如权利要求4所述的非绩优人员培训方法，其特征在于，所述判断所述模型预测准确率是否小于预设阈值的步骤之后还包括：
    若所述模型预测准确率小于所述预设阈值，调整所述神经网络模型的参数，并利用所述训练集重新对调整后的神经网络模型进行训练；
    利用所述验证集对重新训练的神经网络模型进行验证，并根据每一验证结果重新统计得到一模型预测准确率，并判断重新统计得到的模型预测准确率是否小于预设阈值；
    若所述重新统计得到的模型预测准确率不小于所述预设阈值，将所述重新训练得到的神经网络模型作为所述潜力绩优识别模型；及
    若所述重新统计得到的模型预测准确率小于所述预设阈值，重复上述步骤直至通过所述验证集验证得到的模型预测准确率不小于所述预设阈值；
    其中，所述神经网络模型的参数包括总层数、每一层的神经元数。
  6. 如权利要求1所述的非绩优人员培训方法，其特征在于，所述计算所述潜力绩优类型中的绩优人员的每一行为因子的均值与标准差的步骤包括：
    获取所述潜力绩优类型中包含的每一绩优人员的行为因子及行为因子值；及
    根据每一绩优人员的一行为因子的行为因子值计算得出该行为因子的均值与标准差。
  7. 如权利要求1所述的非绩优人员培训方法，其特征在于，所述将所述非绩优人员的每一行为因子值与所述潜力绩优类型的每一行为因子的均值进行一一对应比较的步骤之后还包括：
    当所述非绩优人员的一行为因子值与对应的行为因子的均值的差值不大于标准差的预设倍数时，判定所述非绩优人员不需要选择提升该行为因子的培训课程。
  8. 一种非绩优人员培训系统，其特征在于，所述系统包括：
    获取模块，用于获取多个绩优人员的样本数据，其中所述样本数据包括行为因子数据及非行为因子数据，多个所述绩优人员分属于多种绩优类型，每一所述绩优人员分属于一种绩优类型；
    建立模块，用于根据多个所述绩优人员的非行为因子数据建立并训练得到一潜力绩优识别模型；
    第一计算模块，用于将非绩优人员的非行为因子数据输入至所述潜力绩优识别模型计算得出所述非绩优人员成长为每一绩优类型的概率值；
    选取模块，用于从计算得出的所有概率值中选取一最大概率值，并将最大概率值对应的绩优类型作为所述非绩优人员的潜力绩优类型；
    第二计算模块，用于计算所述潜力绩优类型中的绩优人员的每一行为因子的均值与标准差，并将所述非绩优人员的每一行为因子值与所述潜力绩优类型的每一行为因子的均值进行一一对应比较；及
    判定模块，用于在所述非绩优人员的一行为因子值与对应的行为因子的 均值的差值大于标准差的预设倍数时，判定所述非绩优人员需要选择提升该行为因子的培训课程。
  9. 一种计算机装置，所述计算机装置包括处理器及存储器，所述存储器上存储有若干计算机可读指令，其特征在于，所述处理器用于执行存储器中存储的计算机可读指令时实现以下步骤:
    获取多个绩优人员的样本数据，其中所述样本数据包括行为因子数据及非行为因子数据，多个所述绩优人员分属于多种绩优类型，每一所述绩优人员分属于一种绩优类型；
    根据多个所述绩优人员的非行为因子数据建立并训练得到一潜力绩优识别模型；
    将非绩优人员的非行为因子数据输入至所述潜力绩优识别模型计算得出所述非绩优人员成长为每一绩优类型的概率值；
    从计算得出的所有概率值中选取一最大概率值，并将最大概率值对应的绩优类型作为所述非绩优人员的潜力绩优类型；
    计算所述潜力绩优类型中的绩优人员的每一行为因子的均值与标准差，并将所述非绩优人员的每一行为因子值与所述潜力绩优类型的每一行为因子的均值进行一一对应比较；及
    当所述非绩优人员的一行为因子值与对应的行为因子的均值的差值大于标准差的预设倍数时，判定所述非绩优人员需要选择提升该行为因子的培训课程。
  10. 如权利要求9所述的计算机装置，其特征在于，在所述处理器根据多个所述绩优人员的非行为因子数据建立并训练得到一潜力绩优识别模型时，执行所述计算机可读指令以实现以下步骤：
    建立一神经网络模型，所述神经网络模型包括输入层、多个隐藏层及输出层；及
    利用多个所述绩优人员的非行为因子数据对所述神经网络模型进行训练得到所述潜力绩优识别模型。
  11. 如权利要求9所述的计算机装置，其特征在于，在所述处理器根据多个所述绩优人员的非行为因子数据建立并训练得到一潜力绩优识别模型时，执行所述计算机可读指令以实现以下步骤：
    将多个所述绩优人员的非行为因子数据划分为训练集及验证集；
    建立一神经网络模型，并利用所述训练集对所述神经网络模型进行训练；
    利用所述验证集对训练后的神经网络模型进行验证，并根据每一验证结果统计得到一模型预测准确率；
    判断所述模型预测准确率是否小于预设阈值；
    若所述模型预测准确率不小于所述预设阈值，将训练完成的所述神经网络模型作为所述潜力绩优识别模型。
  12. 如权利要求11所述的计算机装置，其特征在于，在所述判断所述模型预测准确率是否小于预设阈值的步骤之后，所述处理器执行所述计算机可 读指令还用以实现以下步骤：
    若所述模型预测准确率小于所述预设阈值，调整所述神经网络模型的参数，并利用所述训练集重新对调整后的神经网络模型进行训练；
    利用所述验证集对重新训练的神经网络模型进行验证，并根据每一验证结果重新统计得到一模型预测准确率，并判断重新统计得到的模型预测准确率是否小于预设阈值；
    若所述重新统计得到的模型预测准确率不小于所述预设阈值，将所述重新训练得到的神经网络模型作为所述潜力绩优识别模型；及
    若所述重新统计得到的模型预测准确率小于所述预设阈值，重复上述步骤直至通过所述验证集验证得到的模型预测准确率不小于所述预设阈值；
    其中，所述神经网络模型的参数包括总层数、每一层的神经元数。
  13. 如权利要求9所述的计算机装置，其特征在于，在所述处理器计算所述潜力绩优类型中的绩优人员的每一行为因子的均值与标准差时，执行所述计算机可读指令以实现以下步骤：
    获取所述潜力绩优类型中包含的每一绩优人员的行为因子及行为因子值；及
    根据每一绩优人员的一行为因子的行为因子值计算得出该行为因子的均值与标准差。
  14. 如权利要求9所述的计算机装置，其特征在于，所述将所述非绩优人员的每一行为因子值与所述潜力绩优类型的每一行为因子的均值进行一一对应比较的步骤之后，所述处理器执行所述计算机可读指令还用以实现以下步骤：
    当所述非绩优人员的一行为因子值与对应的行为因子的均值的差值不大于标准差的预设倍数时，判定所述非绩优人员不需要选择提升该行为因子的培训课程。
  15. 一种非易失性可读存储介质，其上存储有计算机可读指令，其特征在于，所述计算机可读指令被处理器执行时实现以下步骤：
    获取多个绩优人员的样本数据，其中所述样本数据包括行为因子数据及非行为因子数据，多个所述绩优人员分属于多种绩优类型，每一所述绩优人员分属于一种绩优类型；
    根据多个所述绩优人员的非行为因子数据建立并训练得到一潜力绩优识别模型；
    将非绩优人员的非行为因子数据输入至所述潜力绩优识别模型计算得出所述非绩优人员成长为每一绩优类型的概率值；
    从计算得出的所有概率值中选取一最大概率值，并将最大概率值对应的绩优类型作为所述非绩优人员的潜力绩优类型；
    计算所述潜力绩优类型中的绩优人员的每一行为因子的均值与标准差，并将所述非绩优人员的每一行为因子值与所述潜力绩优类型的每一行为因子的均值进行一一对应比较；及
    当所述非绩优人员的一行为因子值与对应的行为因子的均值的差值大于标准差的预设倍数时，判定所述非绩优人员需要选择提升该行为因子的培训课程。
  16. 如权利要求15所述的存储介质，其特征在于，在所述根据多个所述绩优人员的非行为因子数据建立并训练得到一潜力绩优识别模型时，所述计算机可读指令被所述处理器执行以实现以下步骤：
    建立一神经网络模型，所述神经网络模型包括输入层、多个隐藏层及输出层；及
    利用多个所述绩优人员的非行为因子数据对所述神经网络模型进行训练得到所述潜力绩优识别模型。
  17. 如权利要求15所述的存储介质，其特征在于，在所述根据多个所述绩优人员的非行为因子数据建立并训练得到一潜力绩优识别模型时，所述计算机可读指令被所述处理器执行以实现以下步骤：
    将多个所述绩优人员的非行为因子数据划分为训练集及验证集；
    建立一神经网络模型，并利用所述训练集对所述神经网络模型进行训练；
    利用所述验证集对训练后的神经网络模型进行验证，并根据每一验证结果统计得到一模型预测准确率；
    判断所述模型预测准确率是否小于预设阈值；
    若所述模型预测准确率不小于所述预设阈值，将训练完成的所述神经网络模型作为所述潜力绩优识别模型。
  18. 如权利要求17所述的存储介质，其特征在于，在所述判断所述模型预测准确率是否小于预设阈值的步骤之后，所述计算机可读指令被所述处理器执行还用以实现以下步骤：
    若所述模型预测准确率小于所述预设阈值，调整所述神经网络模型的参数，并利用所述训练集重新对调整后的神经网络模型进行训练；
    利用所述验证集对重新训练的神经网络模型进行验证，并根据每一验证结果重新统计得到一模型预测准确率，并判断重新统计得到的模型预测准确率是否小于预设阈值；
    若所述重新统计得到的模型预测准确率不小于所述预设阈值，将所述重新训练得到的神经网络模型作为所述潜力绩优识别模型；及
    若所述重新统计得到的模型预测准确率小于所述预设阈值，重复上述步骤直至通过所述验证集验证得到的模型预测准确率不小于所述预设阈值；
    其中，所述神经网络模型的参数包括总层数、每一层的神经元数。
  19. 如权利要求15所述的存储介质，其特征在于，在所述计算所述潜力绩优类型中的绩优人员的每一行为因子的均值与标准差时，所述计算机可读指令被所述处理器执行以实现以下步骤：
    获取所述潜力绩优类型中包含的每一绩优人员的行为因子及行为因子值；及
    根据每一绩优人员的一行为因子的行为因子值计算得出该行为因子的均 值与标准差。
  20. 如权利要求15所述的存储介质，其特征在于，所述将所述非绩优人员的每一行为因子值与所述潜力绩优类型的每一行为因子的均值进行一一对应比较的步骤之后，所述计算机可读指令被所述处理器执行还用以实现以下步骤：
    当所述非绩优人员的一行为因子值与对应的行为因子的均值的差值不大于标准差的预设倍数时，判定所述非绩优人员不需要选择提升该行为因子的培训课程。
PCT/CN2019/077511 2018-08-27 2019-03-08 非绩优人员培训方法、系统、计算机装置及存储介质 WO2020042584A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201810983357.4A CN109214448B (zh) 2018-08-27 2018-08-27 非绩优人员培训方法、系统、终端及计算机可读存储介质
CN201810983357.4 2018-08-27

Publications (1)

Publication Number Publication Date
WO2020042584A1 true WO2020042584A1 (zh) 2020-03-05

Family

ID=64985452

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2019/077511 WO2020042584A1 (zh) 2018-08-27 2019-03-08 非绩优人员培训方法、系统、计算机装置及存储介质

Country Status (2)

Country Link
CN (1) CN109214448B (zh)
WO (1) WO2020042584A1 (zh)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112529324A (zh) * 2020-12-18 2021-03-19 平安科技（深圳）有限公司 基于深度学习的产能计算方法、装置、设备及存储介质
CN114997263A (zh) * 2022-04-20 2022-09-02 平安科技（深圳）有限公司 基于机器学习的结训率分析方法、装置、设备及存储介质

Families Citing this family (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109214448B (zh) * 2018-08-27 2023-04-07 平安科技（深圳）有限公司 非绩优人员培训方法、系统、终端及计算机可读存储介质
CN109919811B (zh) * 2019-01-25 2023-09-15 平安科技（深圳）有限公司 基于大数据的保险代理人培养方案生成方法及相关设备
CN112349168A (zh) * 2020-11-10 2021-02-09 国网天津静海供电有限公司 电力调控员沟通协调仿真培训系统及方法

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106126597A (zh) * 2016-06-20 2016-11-16 乐视控股（北京）有限公司 用户属性预测方法及装置
CN108062657A (zh) * 2017-11-30 2018-05-22 朱学松 人才招聘面试方法及系统
CN108280542A (zh) * 2018-01-15 2018-07-13 深圳市和讯华谷信息技术有限公司 一种用户画像模型的优化方法、介质以及设备
CN109214448A (zh) * 2018-08-27 2019-01-15 平安科技（深圳）有限公司 非绩优人员培训方法、系统、终端及计算机可读存储介质

Family Cites Families (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107909272A (zh) * 2017-11-15 2018-04-13 平安科技（深圳）有限公司 员工培训报名方法、应用服务器及计算机可读存储介质

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106126597A (zh) * 2016-06-20 2016-11-16 乐视控股（北京）有限公司 用户属性预测方法及装置
CN108062657A (zh) * 2017-11-30 2018-05-22 朱学松 人才招聘面试方法及系统
CN108280542A (zh) * 2018-01-15 2018-07-13 深圳市和讯华谷信息技术有限公司 一种用户画像模型的优化方法、介质以及设备
CN109214448A (zh) * 2018-08-27 2019-01-15 平安科技（深圳）有限公司 非绩优人员培训方法、系统、终端及计算机可读存储介质

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112529324A (zh) * 2020-12-18 2021-03-19 平安科技（深圳）有限公司 基于深度学习的产能计算方法、装置、设备及存储介质
CN114997263A (zh) * 2022-04-20 2022-09-02 平安科技（深圳）有限公司 基于机器学习的结训率分析方法、装置、设备及存储介质
CN114997263B (zh) * 2022-04-20 2024-05-07 平安科技（深圳）有限公司 基于机器学习的结训率分析方法、装置、设备及存储介质

Also Published As

Publication number Publication date
CN109214448A (zh) 2019-01-15
CN109214448B (zh) 2023-04-07

Similar Documents

Publication Publication Date Title
WO2020042584A1 (zh) 非绩优人员培训方法、系统、计算机装置及存储介质
WO2020042583A1 (zh) 潜力绩优人员类型识别方法、系统、计算机装置及介质
WO2020119272A1 (zh) 风险识别模型训练方法、装置及服务器
CN110070391B (zh) 数据处理方法、装置、计算机可读介质及电子设备
CN111967971B (zh) 银行客户数据处理方法及装置
WO2021190086A1 (zh) 面审风险控制方法、装置、计算机设备及存储介质
JP2017522676A (ja) 才能のデータ駆動型識別システム及び方法
Sundar A comparative study for predicting students academic performance using Bayesian network classifiers
KR102285665B1 (ko) 학습 커리큘럼 제공 방법, 시스템 및 장치
WO2020155300A1 (zh) 一种模型预测方法及装置
Martín-Morató et al. What is the ground truth? reliability of multi-annotator data for audio tagging
US10930401B2 (en) Pet evaluation and triage system
CN112418059B (zh) 一种情绪识别的方法、装置、计算机设备及存储介质
WO2021042842A1 (zh) 基于ai面试系统的面试方法、装置和计算机设备
Kaibel et al. Rethinking the gold standard with multi-armed bandits: machine learning allocation algorithms for experiments
Karimi-Haghighi et al. Predicting early dropout: Calibration and algorithmic fairness considerations
KR20140080773A (ko) 다중 의사 결정 문제 해결을 위한 구간 회색수 및 엔트로피 기반 해법 연산 방법
JP2020160551A (ja) 人事項目の分析支援装置、分析支援方法、プログラム、および記録媒体
CN111368131B (zh) 用户关系识别方法、装置、电子设备及存储介质
CN112115908A (zh) 一种社交能力评估方法及其装置
Larsen A framework for understanding AI-induced field change: How AI technologies are legitimized and institutionalized
CN110782119A (zh) 基于人工智能的保险代理人甄选方法、装置及设备
CN111353728A (zh) 一种风险分析方法和系统
JP7058368B1 (ja) 処理装置及び処理方法
Schulwitz et al. Evaluating citizen science outreach: A case-study with The Peregrine Fund’s American Kestrel Partnership

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 19853649

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 19853649

Country of ref document: EP

Kind code of ref document: A1