Model training method, task execution result determination method, device and equipment

Abstract

The application provides a model training method, a task execution result determination method, a device and equipment, comprising: determining a target task type, and determining a plurality of historical tasks according to the target task type; obtaining a task execution result of each historical task and a plurality of parameter values corresponding to a plurality of parameter types; determining a plurality of target parameter types in the plurality of parameter types according to the task execution result of each historical task and the plurality of parameter values corresponding to the plurality of parameter types; obtaining a plurality of training samples according to the plurality of target parameter types, wherein the training samples include a plurality of sample task parameters of a sample task and a sample execution result of the sample task, the task type of the sample task is the target task type, and the parameter type of the sample task parameter is the target parameter type; training a to-be-trained model according to the plurality of training samples to obtain a target model, and the target model is used for predicting a task execution result of the target task. The efficiency of training the model is improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of computer, and particularly relates to a model training method, a task execution result determination method, device and equipment. BACKGROUND

[0002] In actual work, a work task is usually needed to be risk controlled to ensure the success of task execution.

[0003] In the related art, the execution result of a task can be predicted by a model. In the training process of the model, sample data can be obtained, the sample data can include a plurality of tasks of different types and a plurality of parameter values of each task, the model is trained by the sample data, and a target model is obtained. However, in the above process, because the types of tasks are various, and the parameter types corresponding to the plurality of parameter values of each task can be different, the efficiency of training the model is low. SUMMARY

[0004] The present application provides a model training method, a task execution result determination method, device and equipment to improve the efficiency of training the model.

[0005] In a first aspect, the present application provides a model training method, comprising:

[0006] determining a target task type, and determining a plurality of historical tasks according to the target task type, the task type of the historical task being the target task type;

[0007] obtaining a task execution result of each historical task and a plurality of parameter values corresponding to a plurality of parameter types;

[0008] determining a plurality of target parameter types in the plurality of parameter types according to the task execution result of each historical task and the plurality of parameter values corresponding to the plurality of parameter types, the correlation degree of the target parameter type and the task execution result being greater than or equal to a preset threshold;

[0009] obtaining a plurality of training samples according to the plurality of target parameter types, the training sample including a plurality of sample task parameters of a sample task and a sample execution result of the sample task, the task type of the sample task being the target task type, and the parameter type of the sample task parameter being the target parameter type;

[0010] training a to-be-trained model according to the plurality of training samples to obtain a target model, the target model being used to predict the task execution result of the target task.

[0011] In a possible implementation, the plurality of target parameter types are determined from the plurality of parameter types according to the task execution result of each historical task and the plurality of parameter values corresponding to the plurality of parameter types, including:

[0012] The plurality of historical tasks are classified according to the task execution result of each historical task, to obtain a plurality of historical task sets, the task execution result of each historical task in the historical task set being the same;

[0013] For any one historical task set, target parameter types corresponding to the historical task set are determined from the plurality of parameter types according to the plurality of parameter values of each historical task in the historical task set.

[0014] The plurality of target parameter types include the target parameter types corresponding to each historical task set.

[0015] In a possible implementation, the target parameter types corresponding to the historical task set are determined from the plurality of parameter types according to the plurality of parameter values of each historical task in the historical task set, including:

[0016] The plurality of parameter values of each historical task in the historical task set are normalized to obtain a plurality of normalized parameter values;

[0017] Parameter distribution information corresponding to each parameter type is determined according to the plurality of normalized parameter values of each historical task in the historical task set;

[0018] The target parameter types corresponding to the historical task set are determined from the plurality of parameter types according to the parameter distribution information corresponding to each parameter type.

[0019] In a possible implementation, the parameter distribution information is a parameter curve diagram; the target parameter types corresponding to the historical task set are determined from the plurality of parameter types according to the parameter distribution information corresponding to each parameter type, including:

[0020] For any one parameter type, a parameter fluctuation value corresponding to the parameter type is determined according to the parameter curve diagram corresponding to the parameter type;

[0021] A parameter type with a parameter fluctuation value less than or equal to a preset fluctuation value is determined as the target parameter type corresponding to the historical task set.

[0022] In a possible implementation, the target model is obtained by training the to-be-trained model according to the plurality of training samples, including:

[0023] The plurality of training samples are divided into a training set and a verification set;

[0024] model training is performed on the to-be-trained model through the training set, to obtain an intermediate model;

[0025] verification processing is performed on the intermediate model according to the verification set, to obtain a processing accuracy of the intermediate model;

[0026] If the processing accuracy is greater than or equal to an accuracy threshold, the intermediate model is determined as the target model.

[0027] If the processing accuracy is less than the accuracy threshold, the intermediate model is updated according to the processing accuracy, to obtain the target model.

[0028] In a possible implementation, the method further includes:

[0029] determining a weight value of each target parameter type in the target model, the weight value being used to indicate an influence degree on the task execution result;

[0030] displaying the weight value corresponding to each target parameter type.

[0031] In a second aspect, the present application provides a method for determining a task execution result, including:

[0032] determining a plurality of target parameter types corresponding to a target task;

[0033] acquiring a plurality of task parameters of the target task according to the plurality of target parameter types;

[0034] processing the plurality of task parameters through a target model to obtain a task execution result of the target task, the target model being determined according to the method of any one of the first aspect.

[0035] In a third aspect, the present application provides a model training device, including a first determining module, a first acquiring module, a second determining module, a second acquiring module and a training module, wherein,

[0036] The first determining module is used to determine a target task type, and determine a plurality of historical tasks according to the target task type, the task type of the historical task being the target task type;

[0037] The first acquiring module is used to acquire a task execution result of each historical task, and a plurality of parameter values corresponding to a plurality of parameter types;

[0038] The second determining module is used to determine a plurality of target parameter types in the plurality of parameter types according to the task execution result of each historical task and the plurality of parameter values corresponding to the plurality of parameter types, the correlation degree of the target parameter type with the task execution result being greater than or equal to a preset threshold;

[0039] The second obtaining module is configured to obtain a plurality of training samples according to the plurality of target parameter types, wherein the training samples include a plurality of sample task parameters of a sample task and a sample execution result of the sample task, the task type of the sample task is the target task type, and the parameter type of the sample task parameter is the target parameter type;

[0040] The training module is configured to train a to-be-trained model according to the plurality of training samples to obtain a target model, wherein the target model is configured to predict a task execution result of the target task.

[0041] In a possible implementation, the second determining module is specifically configured to:

[0042] perform classification processing on the plurality of historical tasks according to the task execution results of the historical tasks to obtain a plurality of historical task sets, wherein the historical tasks in the historical task set have the same task execution result;

[0043] For any one historical task set, determine the target parameter type corresponding to the historical task set from the plurality of parameter types according to the plurality of parameter values of the historical tasks in the historical task set;

[0044] The plurality of target parameter types include the target parameter type corresponding to each historical task set.

[0045] In a possible implementation, the second determining module is specifically configured to:

[0046] perform normalization processing on the plurality of parameter values of the historical tasks in the historical task set to obtain a plurality of normalized parameter values;

[0047] determine parameter distribution information corresponding to each parameter type according to the plurality of normalized parameter values of the historical tasks in the historical task set;

[0048] determine the target parameter type corresponding to the historical task set from the plurality of parameter types according to the parameter distribution information corresponding to each parameter type.

[0049] In a possible implementation, the parameter distribution information is a parameter curve diagram; and the second determining module is specifically configured to:

[0050] For any one parameter type, determine a parameter fluctuation value corresponding to the parameter type according to the parameter curve diagram corresponding to the parameter type;

[0051] determine, as the target parameter type corresponding to the historical task set, a parameter type whose parameter fluctuation value is less than or equal to a preset fluctuation value.

[0052] In a possible implementation, the training module is specifically configured to:

[0053] divide the plurality of training samples into a training set and a validation set;

[0054] perform model training on the to-be-trained model through the training set to obtain an intermediate model;

[0055] perform validation processing on the intermediate model according to the validation set to obtain a processing accuracy of the intermediate model;

[0056] if the processing accuracy is greater than or equal to an accuracy threshold, the intermediate model is determined as the target model;

[0057] if the processing accuracy is less than the accuracy threshold, the intermediate model is updated according to the processing accuracy to obtain the target model.

[0058] In a possible implementation, the apparatus further includes a third determination module and a display module:

[0059] the third determination module is configured to determine a weight value of each target parameter type in the target model, the weight value being used to indicate an influence degree on the task execution result;

[0060] the display module is configured to display the weight value corresponding to each target parameter type.

[0061] In a fourth aspect, an embodiment of the present application provides a determination apparatus of a task execution result, including a determination module, an acquisition module and a processing module, wherein,

[0062] the determination module is configured to determine a plurality of target parameter types corresponding to a target task;

[0063] the acquisition module is configured to acquire a plurality of task parameters of the target task according to the plurality of target parameter types;

[0064] the processing module is configured to process the plurality of task parameters through a target model to obtain a task execution result of the target task, the target model being determined according to the method in any one of the first aspect.

[0065] In a fifth aspect, the present application provides an electronic device, including a processor and a memory connected with the processor in communication;

[0066] the memory stores computer execution instructions;

[0067] the processor executes the computer execution instructions stored in the memory to implement the method in any one of the first aspect.

[0068] In a sixth aspect, the present application provides an electronic device, comprising: a processor, and a memory connected with the processor in communication;

[0069] The memory stores computer-executable instructions.

[0070] The processor executes the computer-executable instructions stored in the memory to implement the method according to any one of the second aspect.

[0071] In a seventh aspect, the present application provides a computer-readable storage medium, wherein the computer-readable storage medium stores computer-executable instructions, and the computer-executable instructions are executed by a processor to implement the method according to any one of the first aspect.

[0072] In an eighth aspect, the present application provides a computer-readable storage medium, wherein the computer-readable storage medium stores computer-executable instructions, and the computer-executable instructions are executed by a processor to implement the method according to any one of the second aspect.

[0073] In a ninth aspect, the present application provides a computer program product, comprising a computer program, and the computer program is executed by a processor to implement the method according to any one of the first aspect.

[0074] In a tenth aspect, the present application provides a computer program product, comprising a computer program, and the computer program is executed by a processor to implement the method according to any one of the second aspect.

[0075] In the embodiments of the present application, a model training method and a task execution result determination method and device are provided. The electronic device can determine a target task type, and determine a plurality of historical tasks according to the target task type. The electronic device can obtain a task execution result of each historical task and a plurality of parameter values corresponding to a plurality of parameter types, and then determine a plurality of target parameter types from the plurality of parameter types according to the task execution result of each historical task and the plurality of parameter values corresponding to the plurality of parameter types, and obtain a plurality of training samples according to the plurality of target parameter types, and train a to-be-trained model according to the plurality of training samples to obtain a target model. Since the electronic device can determine a target task type, determine a plurality of historical tasks according to the target task type, and determine a plurality of target parameter types from a plurality of parameter types, and then obtain a plurality of training samples according to the target parameter types, a large number of useless parameters are avoided, and the efficiency of training the model is improved. BRIEF DESCRIPTION OF DRAWINGS

[0076] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present application and serve to explain the principles of the present application together with the specification.

[0077] Figure 1A scenario schematic diagram provided for an exemplary embodiment of the present application;

[0078] Figure 2 A flow schematic diagram of a model training method provided for an embodiment of the present application;

[0079] Figure 3 A flow schematic diagram of another model training method provided for an embodiment of the present application;

[0080] Figure 4 A process schematic diagram of a model training process provided for an embodiment of the present application;

[0081] Figure 5 A flow schematic diagram of a task execution result determination method provided for an embodiment of the present application;

[0082] Figure 6 A structure schematic diagram of a model training apparatus provided for an embodiment of the present application;

[0083] Figure 7 A structure schematic diagram of a task execution result determination apparatus provided for an embodiment of the present application;

[0084] Figure 8 A structure schematic diagram of an electronic device provided for an embodiment of the present application.

[0085] Through the above-described drawings, the explicit embodiments of the present application have been shown, and will be described in more detail hereinafter. These drawings and textual descriptions are not intended to limit the scope of the concept of the present application by any means, but to illustrate the concept of the present application to those skilled in the art by referring to specific embodiments. DETAILED DESCRIPTION

[0086] The exemplary embodiments will be described in detail herein with reference to the attached drawings. In the following description, the same numbers are used to indicate the same or similar components. The embodiments described in the following exemplary embodiments are not intended to represent all embodiments consistent with the present application. Rather, they are merely examples of apparatuses and methods consistent with some aspects of the present application as detailed in the appended claims.

[0087] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data for analysis, stored data, displayed data, etc.) involved in the present application are all information and data authorized by the user or authorized by all parties, and the collection, use and processing of related data need to comply with relevant laws, regulations and standards, and provide corresponding operation portal for user to choose authorization or refusal.

[0088] It should be noted that the model training method and device of the present application can be used in the field of computer technology, and can also be used in any field other than the field of computer technology. The application field of the model training method and device of the present application is not limited.

[0089] For ease of understanding, below, in combination with Figure 1 The application scenarios to which the embodiments of the present application are applied are described.

[0090] Figure 1 A scene schematic diagram is provided for the exemplary embodiments of the present application. Please refer to Figure 1 The electronic device can determine a plurality of historical tasks, and obtain a task execution result of each historical task, and a plurality of parameter values corresponding to a plurality of parameter types.

[0091] For example, the electronic device can obtain a plurality of parameter values and a task execution result 1 of a historical task 1, can obtain a plurality of parameter values and a task execution result 2 of a historical task 2, …, and can obtain a plurality of parameter values and a task execution result n (n is an integer greater than or equal to 1) of a historical task n.

[0092] The electronic device can determine a plurality of target parameter types from the plurality of parameter types according to the task execution result of each historical task and the plurality of parameter values corresponding to the plurality of parameter types, and then obtain training samples according to the plurality of target parameter types, and then train the to-be-trained model through the training samples to obtain a target model. For example, if there are 500 historical tasks, the electronic device can determine 7 target parameter types from 20 parameter types according to the task execution result of each historical task and the parameter values corresponding to the 20 parameter types, and then obtain training samples according to the 7 target parameter types. For example, the training samples can include a plurality of sample task parameters and a sample execution result 1 of a sample task 1, a plurality of sample task parameters and a sample execution result 2 of a sample task 2, …, a plurality of sample task parameters and a sample execution result m (m is an integer greater than or equal to 1) of a sample task m.

[0093] After the electronic device obtains the training samples, the electronic device can train the to-be-trained model through the training samples to obtain a target model. The target model can be used to predict the task execution result of a target task.

[0094] In the related art, the execution result of a task can be predicted by a model. In the training process of the model, sample data can be obtained, the sample data can include a plurality of tasks of different types and a plurality of parameter values of each task, the model is trained through the sample data to obtain a target model. However, in the above process, due to the variety of tasks, and the parameter types corresponding to the plurality of parameter values of each task may be different, resulting in low efficiency of training the model.

[0095] In the embodiment of the present application, the electronic device can determine the target task type, and determine a plurality of historical tasks according to the target task type. The electronic device can obtain the task execution result of each historical task, and a plurality of parameter values corresponding to a plurality of parameter types, and then determine a plurality of target parameter types from the plurality of parameter types according to the task execution result of each historical task and the plurality of parameter values corresponding to the plurality of parameter types, and obtain a plurality of training samples according to the plurality of target parameter types, and train the to-be-trained model according to the plurality of training samples to obtain a target model. Since the electronic device can determine the target task type, and determine a plurality of historical tasks according to the target task type, and can determine a plurality of target parameter types from the plurality of parameter types, and then obtain a plurality of training samples according to the target parameter types, a large number of useless parameters are avoided, and the efficiency of training the model is improved.

[0096] The technical solutions of the present application and how the technical solutions of the present application solve the above technical problems will be described in detail below with specific embodiments. The following specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described again in some embodiments. The embodiments of the present application will be described below with reference to the accompanying drawings.

[0097] Figure 2 A flowchart of a model training method provided by an embodiment of the present application is shown. Please refer to Figure 2 The method can include:

[0098] S201, determine a target task type, and determine a plurality of historical tasks according to the target task type.

[0099] The execution subject of the embodiment of the present application can be an electronic device, or a model training device provided in the electronic device. The model training device can be implemented by software, or by a combination of software and hardware. The model training device can be a processor in the electronic device. For ease of understanding, in the following, the execution subject is taken as an example to be described as an electronic device.

[0100] For any one task, the task can have a corresponding task type. For example, if task 1 is a software development task, the corresponding task type can be a software development type. If task 2 is a product development task, the corresponding task type can be a product development type.

[0101] Optionally, the electronic device can store a plurality of tasks. The electronic device can determine a plurality of historical tasks from the plurality of tasks according to a target task type. The historical tasks can have the target task type. For example, the electronic device can store 1000 tasks. If the target task type is software development, the electronic device can determine 200 historical tasks with the software development type from the 1000 tasks.

[0102] S202, obtaining a task execution result of each historical task and a plurality of parameter values corresponding to a plurality of parameter types.

[0103] The task execution result can be different for different types of historical tasks. For example, if the historical task is a software development task, the task execution result can include success or failure. If the historical task is a product development task, the task execution result can include early completion, on-time completion, and overdue completion.

[0104] For any historical task, the historical task has a corresponding task execution result and a plurality of parameter values corresponding to a plurality of parameter types. For example, for historical task 1, the execution result of historical task 1 can be success, and historical task 1 can have five parameter values, which are: parameter value 1 corresponding to task period, parameter value 2 corresponding to task budget, parameter value 3 corresponding to the number of performers, parameter value 4 corresponding to cost, and parameter value 5 corresponding to task level.

[0105] Optionally, the electronic device can obtain the task execution result of each historical task and the plurality of parameter values corresponding to the plurality of parameter types in the database. For example, if there are 200 historical tasks and 20 parameter types, the electronic device can obtain the task execution result of each historical task and the parameter values corresponding to the 20 parameter types, respectively.

[0106] S203, determining a plurality of target parameter types from the plurality of parameter types according to the task execution result of each historical task and the plurality of parameter values corresponding to the plurality of parameter types.

[0107] Optionally, the relevance of the target parameter type to the task execution result can be greater than or equal to a preset threshold, that is, the target parameter type is a key parameter type that affects the task execution result. For example, the preset threshold can be 80%, and the target parameter type can be the task budget, and the relevance to the task execution result can be greater than 80%.

[0108] In an optional embodiment, the target parameter types can be determined from the task execution results of each historical task and the parameter values corresponding to the parameter types in the following manner: the historical tasks are classified according to the task execution results of each historical task to obtain a plurality of historical task sets; for any one historical task set, the target parameter types corresponding to the historical task set are determined from the parameter values of the historical tasks in the historical task set; and the target parameter types include the target parameter types corresponding to each historical task set.

[0109] Optionally, the electronic device can determine the task execution results of each historical task, and classify the historical tasks according to the task execution results of each historical task to obtain a plurality of historical task sets. The historical tasks in the historical task sets have the same task execution results.

[0110] For example, if there are 200 historical tasks, 170 of which have a successful task execution result and 30 of which have a failed task execution result, the electronic device can classify the 200 historical tasks to obtain two historical task sets, historical task set 1 and historical task set 2. Historical task set 1 can include the 170 historical tasks with a successful task execution result, and historical task set 2 can include the 30 historical tasks with a failed task execution result.

[0111] Optionally, for any one historical task set, the electronic device can determine the target parameter types corresponding to the historical task set from the parameter values of the historical tasks in the historical task set. For example, if historical task set 1 includes 170 historical tasks, each of which has a plurality of parameter values corresponding to 20 parameter types, the electronic device can determine the target parameter types corresponding to historical task set 1 from the parameter values of each historical task in the 20 parameter types. Assume that the target parameter types corresponding to historical task set 1 are four, namely target parameter type 1, target parameter type 2, target parameter type 3, and target parameter type 4.

[0112] Optionally, the electronic device can determine that the target parameter types include the target parameter types corresponding to each historical task set. For example, if the target parameter types corresponding to historical task set 1 are four, namely target parameter type 1, target parameter type 2, target parameter type 3, and target parameter type 4, and the target parameter types corresponding to historical task set 2 are three, namely target parameter type 5, target parameter type 6, and target parameter type 7, the electronic device can determine that the target parameter types are seven, namely target parameter type 1, target parameter type 2, …, and target parameter type 7.

[0113] S204, obtain a plurality of training samples according to a plurality of target parameter types.

[0114] For any one training sample, the training sample can include a plurality of sample task parameters of a sample task and a sample execution result of the sample task.

[0115] The task type of the sample task can be a target task type, and the parameter type of the sample task parameter can be a target parameter type.

[0116] For example, if there are 7 target parameter types, the electronic device can obtain the sample task parameters corresponding to the 7 target parameter types in the sample task 1 respectively according to the 7 target parameter types, and the sample execution result of the sample task 1, to obtain the training sample 1. The task type of the sample task 1 can be a software development type.

[0117] Optionally, the electronic device can obtain a plurality of training samples according to a plurality of target parameter types. Each training sample can include a plurality of sample task parameters of a sample task and a sample execution result of the sample task. For example, if there are 7 target parameter types, the electronic device can obtain 180 training samples according to the 7 target parameter types, which can be respectively: a plurality of sample task parameters and a sample execution result corresponding to the sample task 1, a plurality of sample task parameters and a sample execution result corresponding to the sample task 2,..., a plurality of sample task parameters and a sample execution result corresponding to the sample task 180. The task type of the 180 sample tasks can be a software development type. The parameter type of the plurality of sample task parameters corresponding to each sample task is a target parameter type.

[0118] S205, training the to-be-trained model according to the plurality of training samples to obtain a target model.

[0119] The target model can be used to predict the task execution result of the target task.

[0120] Optionally, the to-be-trained model can be a decision tree model.

[0121] After the electronic device obtains a plurality of training samples, the to-be-trained model can be trained by the plurality of training samples to obtain a target model. For example, if there are 180 training samples, the to-be-trained model can be a decision tree model, and the electronic device can train the decision tree model according to the 180 training samples to obtain the target model.

[0122] In the embodiment of the present application, the electronic device can determine a target task type, and determine a plurality of historical tasks according to the target task type. The electronic device can obtain a task execution result of each historical task, and a plurality of parameter values corresponding to a plurality of parameter types, and then determine a plurality of target parameter types from the plurality of parameter types according to the task execution result of each historical task and the plurality of parameter values corresponding to the plurality of parameter types, and obtain a plurality of training samples according to the plurality of target parameter types, and train the to-be-trained model according to the plurality of training samples to obtain a target model. Since the electronic device can determine a target task type, and determine a plurality of historical tasks according to the target task type, and can determine a plurality of target parameter types from the plurality of parameter types, and then obtain a plurality of training samples according to the target parameter types, a large number of useless parameters are avoided, and the efficiency of training the model is improved.

[0123] In the following, Figure 2 on the basis of the embodiments shown in Figure 3 The model training method is described in detail.

[0124] Figure 3 Another model training method provided in the embodiment of the present application is shown in the flowchart. Please refer to Figure 3 The method can include:

[0125] S301, determine a target task type, and determine a plurality of historical tasks according to the target task type.

[0126] S302, obtain a task execution result of each historical task, and a plurality of parameter values corresponding to a plurality of parameter types.

[0127] It should be noted that the execution process of steps S301-S302 can refer to the execution process of steps S201-S202, which will not be described here.

[0128] S303, according to the task execution result of each historical task, the plurality of historical tasks are classified and processed to obtain a plurality of historical task sets.

[0129] For example, if the task execution result of the historical task includes success and failure, if there are 200 historical tasks, if there are 200 historical tasks, 170 of which have a task execution result of success, and 30 of which have a task execution result of failure, the electronic device can classify and process the 200 historical tasks to obtain 2 historical task sets, namely historical task set 1 and historical task set 2. The historical task set 1 can include 170 historical tasks with successful execution, and the historical task set 2 can include 30 historical tasks with failed execution.

[0130] S304. For any one historical task set, target parameter types corresponding to the historical task set are determined from the plurality of parameter types according to the plurality of parameter values of the historical tasks in the historical task set.

[0131] In an optional embodiment, the target parameter types corresponding to the historical task set can be determined from the plurality of parameter types according to the plurality of parameter values of the historical tasks in the historical task set by: normalizing the plurality of parameter values of the historical tasks in the historical task set to obtain a plurality of normalized parameter values; determining parameter distribution information corresponding to each parameter type according to the plurality of normalized parameter values of the historical tasks in the historical task set; and determining the target parameter types corresponding to the historical task set from the plurality of parameter types according to the parameter distribution information corresponding to each parameter type.

[0132] Optionally, the electronic device can normalize the plurality of parameter values of the historical tasks in the historical task set to obtain a plurality of normalized parameter values.

[0133] Optionally, for any one parameter type, the electronic device can determine parameter values corresponding to the parameter type in each historical task according to the parameter type, and then determine a maximum parameter value, normalize the plurality of parameter values of the parameter type according to the maximum parameter value, and obtain a plurality of normalized parameter values.

[0134] For example, if the historical task set 1 includes 170 historical tasks, and each historical task has 20 parameter types corresponding to parameter values, for any one parameter type, the electronic device can determine parameter value 1 corresponding to the parameter type in historical task 1, parameter value 2 corresponding to the parameter type in historical task 2, …, and parameter value 170 corresponding to the parameter type in historical task 170 according to the parameter type, and normalize the 170 parameter values to obtain 170 normalized parameter values.

[0135] For example, if the parameter type 1 is the task period, the electronic device can determine the task periods of the 170 historical tasks and determine the maximum task period. For example, if the maximum task period is 15 months, the parameter value 1 corresponding to the task period in historical task 1 is 10 months, the parameter value 1 can be normalized to obtain the normalized parameter value 1 as 10 / 15=0.66; if the parameter value 2 corresponding to the task period in historical task 2 is 8 months, the parameter value 2 can be normalized to obtain the normalized parameter value 2 as 8 / 15=0.53; …; if the parameter value 170 corresponding to the task period in historical task 170 is 15 months, the parameter value 170 can be normalized to obtain the normalized parameter value 170 as 15 / 15=1.

[0136] Optionally, for any one parameter type, the electronic device can determine parameter distribution information corresponding to the parameter type according to the plurality of normalized parameter values of the historical tasks in the historical task set. Optionally, the parameter distribution information can be a parameter curve.

[0137] For example, if the parameter type is the task period, and the historical task set 1 includes 170 historical tasks, the electronic device can determine the parameter distribution information 1 corresponding to the parameter type, i.e., the parameter curve Figure 1 .

[0138] In an optional embodiment, the target parameter type corresponding to the historical task set can be determined from the plurality of parameter types according to the parameter distribution information corresponding to each parameter type in the following manner: for any one parameter type, the parameter fluctuation value corresponding to the parameter type is determined according to the parameter curve corresponding to the parameter type; and the parameter type with a parameter fluctuation value less than or equal to a preset fluctuation value is determined as the target parameter type corresponding to the historical task set.

[0139] The parameter fluctuation value can include a positive value and / or a negative value. For example, the parameter fluctuation value can be ±5.

[0140] Optionally, the electronic device can determine the parameter median corresponding to each parameter type according to the parameter curve corresponding to the parameter type, and determine the parameter fluctuation value according to the parameter median, the maximum parameter value, and the minimum parameter value. For example, if the parameter type is the task period, and if the electronic device determines that the parameter median corresponding to the task period is 12 months, the maximum parameter value is 15 months, and the minimum parameter value is 8 months according to the parameter curve corresponding to the task period, the parameter fluctuation value can be determined as ±4 months.

[0141] For any one parameter type, the parameter type can have a corresponding preset fluctuation value.

[0142] Optionally, for any one parameter type, the electronic device can determine the preset fluctuation value corresponding to the parameter type, and determine the parameter type with a parameter fluctuation value less than or equal to the preset fluctuation value as the target parameter type corresponding to the historical task set.

[0143] For example, if the parameter type is the task period, and if the electronic device determines that the preset fluctuation value corresponding to the task period is ±5 months, and if the parameter fluctuation value corresponding to the task period is ±4 months, the task period can be determined as the target parameter type corresponding to the historical task set 1 because the parameter fluctuation value of the task period is less than the preset fluctuation value.

[0144] The electronic device can determine, according to the parameter distribution information corresponding to each parameter type, a target parameter type corresponding to each historical task set in the plurality of parameter types. For example, for historical set 1, the electronic device can determine, according to the parameter distribution information corresponding to each of the 20 parameter types, 4 target parameter types corresponding to historical set 1 in the 20 parameter types, which are target parameter type 1, target parameter type 2, target parameter type 3, and target parameter type 4, respectively. For historical set 2, the electronic device can determine, according to the parameter distribution information corresponding to each of the 20 parameter types, target parameter types corresponding to historical task set 2 in the 20 parameter types, which are target parameter type 5, target parameter type 6, and target parameter type 7, respectively.

[0145] S305, determining that the plurality of target parameter types include a target parameter type corresponding to each historical task set.

[0146] For example, if the target parameter types corresponding to historical task set 1 are 4, which are target parameter type 1, target parameter type 2, target parameter type 3, and target parameter type 4, and the target parameter types corresponding to historical task set 2 are 3, which are target parameter type 5, target parameter type 6, and target parameter type 7, the electronic device can determine that the target parameter types are 7, which are target parameter type 1, target parameter type 2, …, and target parameter type 7.

[0147] S306, obtaining a plurality of training samples according to the plurality of target parameter types.

[0148] It should be noted that the execution process of step S306 can refer to the execution process of step S204, which will not be described here.

[0149] S307, training a to-be-trained model according to the plurality of training samples to obtain a target model.

[0150] In an optional embodiment, the to-be-trained model can be trained according to the plurality of training samples to obtain the target model in the following feasible manner: dividing the plurality of training samples into a training set and a validation set; training the to-be-trained model through the training set to obtain an intermediate model; performing validation processing on the intermediate model according to the validation set to obtain a processing accuracy of the intermediate model; if the processing accuracy is greater than or equal to an accuracy threshold, the intermediate model is determined as the target model; if the processing accuracy is less than the accuracy threshold, the intermediate model is updated according to the processing accuracy to obtain the target model.

[0151] Optionally, the plurality of training samples can be divided into a training set and a validation set according to the principle of mutual exclusion, and the sample tasks in the training set and the validation set do not coincide. For example, if there are 180 training samples, the 180 training samples can be divided into a training set and a validation set. The training set can include 100 training samples, i.e., the plurality of sample task parameters of 100 sample tasks in the training set, and the sample execution results of each sample task. The validation set can include 80 training samples, i.e., the plurality of sample task parameters of 80 sample tasks in the validation set, and the sample execution results of each sample task.

[0152] The electronic device can perform model training on the to-be-trained model through the training set to obtain an intermediate model. For example, if the to-be-trained model is a decision tree model and the training set includes 100 sample tasks, the electronic device can perform model training on the to-be-trained model through the 100 sample tasks in the training set to obtain an intermediate model.

[0153] Optionally, the electronic device can perform validation processing on the intermediate model according to the plurality of sample tasks in the validation set to obtain the processing accuracy of the intermediate model, i.e., the electronic device can process the plurality of sample task parameters of the plurality of sample tasks in the validation set through the intermediate model to obtain the predicted execution results of the plurality of sample tasks, and compare the predicted execution results of the plurality of sample tasks with the sample execution results of the plurality of sample tasks one by one to obtain the processing accuracy.

[0154] For example, if the validation set can include 80 sample tasks, the electronic device can process the plurality of sample task parameters of the 80 sample tasks in the validation set through the intermediate model respectively to obtain the predicted execution results of the 80 sample tasks, and compare the predicted execution results of the 80 sample tasks with the sample task execution results of the 80 sample tasks one by one to obtain the processing accuracy. For example, if 70 predicted execution results are consistent with the corresponding sample task execution results among the 80 predicted execution results, it can be determined that the processing accuracy is 70 / 80=87.5%.

[0155] Optionally, after the electronic device determines the processing accuracy, if the processing accuracy is greater than or equal to the accuracy threshold, it means that the intermediate model has reached the expected requirement, and the intermediate model can be determined as the target model; if the processing accuracy is less than the accuracy threshold, it means that the processing result of the intermediate model is not accurate enough, and the intermediate model can be updated according to the processing accuracy to obtain the target model.

[0156] Optionally, the intermediate model can be updated by adjusting the model parameters or adjusting the iteration algorithm of the model.

[0157] Optionally, the accuracy threshold can be artificially preset. For example, the accuracy threshold can be 95%.

[0158] For example, if the precision threshold is 95%, and the processing accuracy is 99%, the intermediate model can be determined as the target model because the processing accuracy 99% is greater than the precision threshold 95%; if the processing accuracy is 87.5%, the intermediate model can be adjusted in model parameters or iterative algorithm of the model according to the processing accuracy 87.5% for updating processing to obtain the target model because the processing accuracy 87.5% is less than the precision threshold 95%.

[0159] S308, determining a weight value of each target parameter type in the target model.

[0160] The weight value can be used to indicate the influence degree on the task execution result. The greater the weight value, the greater the influence on the task execution result. For example, if the weight value of the target parameter type 1 in the target model is 0.5, and the weight value of the target parameter type 2 in the target model is 0.3, the influence of the target parameter type 1 on the task execution result is greater than that of the target parameter type 2.

[0161] Optionally, the electronic device can determine the weight value of each target parameter type through the target model. For example, if there are 7 target parameter types, the weight values of the 7 target parameter types can be determined through the target model.

[0162] S309, displaying the weight value corresponding to each target parameter type.

[0163] Optionally, the electronic device can display the weight value corresponding to each target parameter type, and can also sort the weight value corresponding to each target parameter type to obtain a sorting result and display the sorting result.

[0164] For example, if the electronic device determines the weight values of 7 target parameter types and sorts as shown in Table 1:

[0165] Table 1

[0166] Target parameter type Weight value Target parameter type 1 0.46 Target parameter type 2 0.21 Target parameter type 3 0.14 Target parameter type 4 0.12 Target parameter type 5 0.07 Target parameter type 6 0.00 Target parameter type 7 0.00

[0167] The electronic device can display the weight value corresponding to each target parameter type and the sorting result.

[0168] Optionally, the electronic device can also send the weight value corresponding to each target parameter type to a preset device.

[0169] Optionally, the target model can be visually analyzed, and an influence relationship diagram of each target parameter type on the task execution result can be drawn; a single decision tree can also be drawn for the target model to intuitively reflect the influence of each target parameter type on the task execution result.

[0170] In the embodiment of the present application, the electronic device can determine the target task type, and determine a plurality of historical tasks according to the target task type. The electronic device can obtain the task execution result of each historical task, and a plurality of parameter values corresponding to a plurality of parameter types, and perform classification processing on the plurality of historical tasks according to the task execution result of each historical task, to obtain a plurality of historical task sets. For any one historical task set, the target parameter type corresponding to the historical task set is determined from the plurality of parameter types according to the plurality of parameter values of the historical tasks in the historical task set. The plurality of target parameter types including the target parameter type corresponding to each historical task set can be determined. The electronic device can obtain a plurality of training samples according to the plurality of target parameter types, and train the to-be-trained model according to the plurality of training samples to obtain a target model, and can also determine the weight value of each target parameter type in the target model, and display the weight value corresponding to each target parameter type. Since the electronic device can determine the target task type, and determine a plurality of historical tasks according to the target task type, and can determine a plurality of target parameter types from a plurality of parameter types, and then obtain a plurality of training samples according to the target parameter types, a large number of useless parameters are avoided, and the efficiency of training the model is improved.

[0171] In the following, on the basis of any of the above embodiments, in combination with Figure 4 , the above model training process is further described in detail.

[0172] Figure 4 A process schematic diagram of a model training process provided by the embodiment of the present application is shown in FIG. 6. Referring to FIG. 6, Figure 4 , the electronic device can determine the target task type, and determine a plurality of historical tasks according to the target task type, and obtain the task execution result of each historical task, and a plurality of parameter values corresponding to a plurality of parameter types, and then determine a plurality of target parameter types according to the task execution result of each historical task, and the plurality of parameter values corresponding to the plurality of parameter types. For example, the electronic device can obtain 200 historical tasks. Each historical task has 20 parameter types, so the electronic device can obtain the task execution result of each historical task, and the parameter values corresponding to the 20 parameter types respectively, and then determine 7 target parameter types, which are target parameter type 1, target parameter type 2, …, target parameter type 7.

[0173] The electronic device can obtain a plurality of training samples according to the plurality of target parameter types. For example, if there are 7 target parameter types, the electronic device can obtain 180 training samples according to the 7 target parameter types, and each training sample can include a plurality of sample task parameters of a sample task, and a sample execution result of the sample task. The task type of the sample task can be the target task type, and the parameter type of the sample task parameter can be the target parameter type.

[0174] The electronic device can divide the plurality of training samples into a training set and a validation set. For example, if there are 180 training samples, the 180 training samples can be divided into a training set and a validation set. The training set can include 100 training samples, i.e., the plurality of sample task parameters of 100 sample tasks and the sample execution results of each sample task. The validation set can include 80 training samples, i.e., the plurality of sample task parameters of 80 sample tasks and the sample execution results of each sample task.

[0175] The electronic device can perform model training on the to-be-trained model through the training set to obtain an intermediate model. For example, if the to-be-trained model is a decision tree model and the training set includes 100 sample tasks, the electronic device can perform model training on the to-be-trained model through the 100 sample tasks in the training set to obtain an intermediate model.

[0176] The electronic device can perform validation processing on the intermediate model through the validation set to obtain the processing accuracy of the intermediate model. If the processing accuracy is greater than or equal to the accuracy threshold, the intermediate model is determined as the target model. If the processing accuracy is less than the accuracy threshold, the intermediate model is updated according to the processing accuracy to obtain the target model. For example, if the validation set can include 80 sample tasks, the electronic device can process the plurality of sample task parameters of the 80 sample tasks in the validation set through the intermediate model to obtain the predicted execution results of the 80 sample tasks, and compare the predicted execution results with the sample task execution results of the 80 sample tasks one by one to obtain the processing accuracy. If the accuracy threshold is 95% and the processing accuracy is 99%, since the processing accuracy 99% is greater than the accuracy threshold 95%, the intermediate model can be determined as the target model. If the processing accuracy is 87.5%, since the processing accuracy 87.5% is less than the accuracy threshold 95%, the intermediate model can be adjusted in model parameters or the iteration algorithm of the model according to the processing accuracy 87.5% to perform update processing to obtain the target model.

[0177] In this embodiment, the electronic device can determine the target task type and identify multiple historical tasks based on the target task type. The electronic device can acquire the task execution result of each historical task and multiple parameter values ​​corresponding to multiple parameter types. Based on the task execution result of each historical task, it can classify the multiple historical tasks to obtain multiple historical task sets. For any historical task set, based on the multiple parameter values ​​of each historical task in the historical task set, it determines the target parameter type corresponding to the historical task set from among the multiple parameter types. It can be determined that the multiple target parameter types include the target parameter type corresponding to each historical task set. The electronic device can acquire multiple training samples based on the multiple target parameter types and train the model to be trained based on the multiple training samples to obtain the target model. It can also determine the weight value of each target parameter type in the target model and display the weight value corresponding to each target parameter type. Since the electronic device can determine the target task type, identify multiple historical tasks based on the target task type, and identify multiple target parameter types from among the multiple parameter types, it can acquire multiple training samples based on the target parameter types, avoiding a large number of useless parameters and improving the efficiency of model training.

[0178] Below, in conjunction with Figure 5 The process of using the model will be explained.

[0179] Figure 5 This is a flowchart illustrating a method for determining a task execution result provided in an embodiment of this application. Please refer to... Figure 5 The method may include:

[0180] S501. Determine the types of multiple target parameters corresponding to the target task.

[0181] Optionally, the target task can be manually input, and the electronic device can determine the types of multiple target parameters corresponding to the target task based on the target task.

[0182] For example, if the target task is to develop software 1, the electronic device can determine multiple target parameter types based on this target task. Assume the electronic device can determine five target parameter types: task duration, task budget, number of personnel, cost, and task level.

[0183] S502. Based on multiple target parameter types, obtain multiple task parameters of the target task.

[0184] Optionally, multiple target parameter types can be manually input as corresponding task parameters, so that the electronic device can determine and obtain multiple task parameters of the target task based on the multiple target parameter types.

[0185] For example, if there are 5 target parameter types, task period, task budget, number of performers, cost, and task level, the electronic device can obtain parameter value 1 corresponding to the task period as 12 months, parameter value 2 corresponding to the task budget as 5 million, parameter value 3 corresponding to the number of performers as 30 people, parameter value 4 corresponding to the cost as 4 million, and parameter value 5 corresponding to the task level as level 3.

[0186] S503, processing the plurality of task parameters through the target model to obtain a task execution result of the target task.

[0187] The target model can be obtained through the model training method.

[0188] For example, if the target task is to develop software 1, and if there are 5 task parameters, task period 12 months, task budget 5 million, number of performers 30 people, cost 4 million, and level 3, the electronic device can process the 5 task parameters through the target model to obtain the task execution result of the target task. It is assumed that the task execution result of the target task is determined to be successful.

[0189] Optionally, the electronic device can also determine the weight values of the plurality of target parameter types through the target model to determine the influence degree of the plurality of target parameter types on the task execution result. The electronic device can also display the weight values of the plurality of target parameter types. The worker can control the target task according to the weight values of the plurality of target parameter types.

[0190] For example, if there are 5 target parameter types corresponding to the task parameters, task period corresponding to parameter value 1 is 12 months, task budget corresponding to parameter value 2 is 5 million, number of performers corresponding to parameter value 3 is 30 people, cost corresponding to parameter value 4 is 4 million, and task level corresponding to parameter value 5 is level 3, the electronic device can determine the weight values corresponding to the 5 target parameter types through the target model. It is assumed that the weight values corresponding to the 5 target parameter types can be 0.46, 0.21, 0.14, 0.12, and 0.07, respectively. The electronic device can display the weight values corresponding to the 5 target parameter types.

[0191] In the embodiments of the present application, the electronic device can determine a plurality of target parameter types corresponding to the target task, and obtain a plurality of task parameters of the target task according to the plurality of target parameter types, and then process the plurality of task parameters through the target model to obtain the task execution result of the target task. Since the target model is obtained through the model training method, the accuracy of determining the task execution result of the target task is improved by determining the task execution result of the target task through the target model.

[0192] Figure 6A structural schematic diagram of a model training device provided by an embodiment of the present application is provided. Please refer to Figure 6 The model training device 10 comprises a first determining module 11, a first obtaining module 12, a second determining module 13, a second obtaining module 14, and a training module 15, wherein,

[0193] The first determining module 11 is configured to determine a target task type, and determine a plurality of historical tasks according to the target task type, wherein the task type of the historical tasks is the target task type.

[0194] The first obtaining module 12 is configured to obtain a task execution result of each historical task, and a plurality of parameter values corresponding to a plurality of parameter types.

[0195] The second determining module 13 is configured to determine a plurality of target parameter types in the plurality of parameter types according to the task execution result of each historical task and the plurality of parameter values corresponding to the plurality of parameter types, wherein the correlation degree between the target parameter types and the task execution result is greater than or equal to a preset threshold.

[0196] The second obtaining module 14 is configured to obtain a plurality of training samples according to the plurality of target parameter types, wherein the training samples comprise a plurality of sample task parameters of a sample task and a sample execution result of the sample task, the task type of the sample task is the target task type, and the parameter type of the sample task parameter is the target parameter type.

[0197] The training module 15 is configured to train a to-be-trained model according to the plurality of training samples to obtain a target model, wherein the target model is used to predict a task execution result of the target task.

[0198] The model training device provided by the embodiment of the present application can execute the technical solutions shown in the above method embodiments, and the implementation principles and beneficial effects are similar, which will not be described here in detail.

[0199] In a possible implementation, the second determining module 13 is specifically configured to:

[0200] According to the task execution result of each historical task, the plurality of historical tasks are classified to obtain a plurality of historical task sets, wherein the task execution results of the historical tasks in the historical task set are the same.

[0201] For any one historical task set, according to the plurality of parameter values of the historical tasks in the historical task set, the historical task set corresponding target parameter types in the plurality of parameter types are determined.

[0202] The plurality of target parameter types comprise the target parameter types corresponding to each historical task set.

[0203] In a possible implementation, the second determining module 13 is specifically configured to:

[0204] normalize the plurality of parameter values of each historical task in the historical task set to obtain a plurality of normalized parameter values;

[0205] determine parameter distribution information corresponding to each parameter type according to the plurality of normalized parameter values of each historical task in the historical task set;

[0206] determine a target parameter type corresponding to the historical task set from the plurality of parameter types according to the parameter distribution information corresponding to each parameter type.

[0207] In a possible implementation, the parameter distribution information is a parameter graph; and the second determining module 13 is specifically configured to:

[0208] for any one parameter type, determine a parameter fluctuation value corresponding to the parameter type according to the parameter graph corresponding to the parameter type;

[0209] determine, as the target parameter type corresponding to the historical task set, a parameter type with a parameter fluctuation value less than or equal to a preset fluctuation value.

[0210] In a possible implementation, the training module 15 is specifically configured to:

[0211] divide the plurality of training samples into a training set and a verification set;

[0212] perform model training on the to-be-trained model through the training set to obtain an intermediate model;

[0213] perform verification processing on the intermediate model according to the verification set to obtain a processing accuracy of the intermediate model;

[0214] if the processing accuracy is greater than or equal to an accuracy threshold, determine the intermediate model as the target model;

[0215] if the processing accuracy is less than the accuracy threshold, perform update processing on the intermediate model according to the processing accuracy to obtain the target model.

[0216] In a possible implementation, the model training apparatus further includes a third determining module 16 and a display module 17,

[0217] the third determining module 16 is configured to determine a weight value of each target parameter type in the target model, the weight value being used to indicate an influence degree on the task execution result;

[0218] The display module 17 is configured to display the weight value corresponding to each target parameter type.

[0219] The model training apparatus provided by the embodiments of the present application can implement the technical solutions shown in the method embodiments, and the implementation principles and beneficial effects are similar, which will not be repeated here.

[0220] Figure 7 A structural diagram of a task execution result determination apparatus is provided in the embodiments of the present application. Please refer to Figure 7 The task execution result determination apparatus 20 comprises a determination module 21, an acquisition module 22 and a processing module 23, wherein,

[0221] The determination module 21 is configured to determine a plurality of target parameter types corresponding to a target task.

[0222] The acquisition module 22 is configured to acquire a plurality of task parameters of the target task according to the plurality of target parameter types.

[0223] The processing module 23 is configured to process the plurality of task parameters by a target model to obtain a task execution result of the target task, wherein the target model is determined according to the model training method.

[0224] The task execution result determination apparatus provided by the embodiments of the present application can implement the technical solutions shown in the method embodiments, and the implementation principles and beneficial effects are similar, which will not be repeated here.

[0225] A structural diagram of an electronic device is provided in the embodiments of the present application. Please refer to Figure 8 The electronic device 30 can comprise a processor 31 and a memory 32. Exemplarily, the processor 31, the memory 32 and each part are connected with each other through a bus 33.

[0226] The memory 32 stores computer execution instructions.

[0227] The processor 31 executes the computer execution instructions stored in the memory 32, so that the processor 31 executes the model training method or the task execution result determination method shown in the above method embodiments.

[0228] The embodiments of the present application provide a computer readable storage medium, wherein the computer readable storage medium stores computer execution instructions, and the computer execution instructions are used to implement the method provided by any of the above embodiments when executed by a processor.

[0229] The embodiments of the present application can also provide a computer program product comprising a computer program, and the computer program can implement the method provided by any of the above embodiments when executed by a processor.

[0230] All or part of the steps of the above-mentioned method embodiments can be completed by program instruction related hardware. The aforementioned program can be stored in a readable memory. The program, when executed, performs steps including the above-mentioned method embodiments; and the aforementioned memory (storage medium) includes: read-only memory (English: read-only memory, abbreviation: ROM), random access memory (English: Random Access Memory, abbreviation: RAM), flash memory, hard disk, solid state disk, magnetic tape (English: magnetic tape), floppy disk (English: floppy disk), optical disc (English: optical disc), and any combination thereof.

[0231] The embodiments of the present application are described with reference to the flowcharts and / or block diagrams according to the methods, devices (systems), and computer program products of the embodiments of the present application. It should be understood that each flow and / or block in the flowcharts and / or block diagrams, and the combination of flows and / or blocks in the flowcharts and / or block diagrams can be implemented by computer program instructions. These computer program instructions can be provided to a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing apparatus to produce a machine, so that the instructions executed by the computer or other programmable data processing apparatus generate a device that implements the functions specified in the flowcharts and / or block diagrams. Figure 1 one or more flows and / or blocks Figure 1 an apparatus that carries out the functions specified in one or more flows and / or blocks.

[0232] These computer program instructions can also be stored in a computer readable memory that can direct the computer or other programmable data processing apparatus to work in a specific manner, so that the instructions stored in the computer readable memory produce a manufactured product including instruction apparatus, which implements the functions specified in the flowcharts and / or block diagrams. Figure 1 one or more flows and / or blocks Figure 1 an apparatus that carries out the functions specified in one or more flows and / or blocks.

[0233] These computer program instructions can also be loaded into a computer or other programmable data processing apparatus, so that a series of operation steps are performed on the computer or other programmable data processing apparatus to produce a computer implemented process, so that the instructions executed on the computer or other programmable data processing apparatus provide a process for implementing the functions specified in the flowcharts and / or block diagrams. Figure 1 one or more flows and / or blocks Figure 1 an apparatus that carries out the functions specified in one or more flows and / or blocks.

[0234] Other embodiments of the application will be apparent to those skilled in the art from consideration of the specification and practice of the application disclosed herein. It is intended that the specification and examples be considered as exemplary only, with the true scope and spirit of the application being indicated by the following claims.

[0235] It is to be understood that the application is not limited to the precise construction herein disclosed and shown in the drawings, and that various modifications and changes can be made by those skilled in the art without departing from the scope of the application. The scope of the application is limited only by the claims that follow.

Claims

1. A model training method, characterized in that, The method comprises the following steps: determining a target task type, and determining a plurality of historical tasks according to the target task type, the task type of the historical tasks being the target task type; the target task type comprises a software development type and a product development type; obtaining a task execution result of each historical task and a plurality of parameter values corresponding to a plurality of parameter types; if the historical task is a software development task, the task execution result comprises success or failure; if the historical task is a product development task, the task execution result comprises early completion, on-time completion and overdue completion; determining a plurality of target parameter types in the plurality of parameter types according to the task execution result of each historical task and the plurality of parameter values corresponding to the plurality of parameter types, the correlation degree of the target parameter type and the task execution result being greater than or equal to a preset threshold; obtaining a plurality of training samples according to the plurality of target parameter types, the training samples comprising a plurality of sample task parameters of a sample task and a sample execution result of the sample task, the task type of the sample task being the target task type, and the parameter type of the sample task parameter being the target parameter type; training a to-be-trained model according to the plurality of training samples to obtain a target model, the target model being used for predicting a task execution result of the target task; determining a plurality of target parameter types in the plurality of parameter types according to the task execution result of each historical task and the plurality of parameter values corresponding to the plurality of parameter types comprises: classifying the plurality of historical tasks according to the task execution result of each historical task to obtain a plurality of historical task sets, the task execution results of the historical tasks in the historical task set being the same; for any one historical task set, determining a target parameter type corresponding to the historical task set in the plurality of parameter types according to the plurality of parameter values of the historical tasks in the historical task set; the plurality of target parameter types comprise the target parameter type corresponding to each historical task set.

2. The method of claim 1, wherein, determining a target parameter type corresponding to the historical task set in the plurality of parameter types according to the plurality of parameter values of the historical tasks in the historical task set comprises: normalizing the plurality of parameter values of the historical tasks in the historical task set to obtain a plurality of normalized parameter values; determining parameter distribution information corresponding to each parameter type according to the plurality of normalized parameter values of the historical tasks in the historical task set; determining a target parameter type corresponding to the historical task set in the plurality of parameter types according to the parameter distribution information corresponding to each parameter type.

3. The method of claim 2, wherein, the parameter distribution information is a parameter curve diagram; determining a target parameter type corresponding to the historical task set in the plurality of parameter types according to the parameter distribution information corresponding to each parameter type comprises: for any one parameter type, determining a parameter fluctuation value corresponding to the parameter type according to the parameter curve diagram corresponding to the parameter type; determining the parameter type with a parameter fluctuation value less than or equal to a preset fluctuation value as the target parameter type corresponding to the historical task set.

4. The method according to any one of claims 1 to 3, characterized in that, The method comprises the following steps: The plurality of training samples are divided into a training set and a verification set; The training set is used to train the to-be-trained model to obtain an intermediate model; The intermediate model is verified according to the verification set to obtain a processing accuracy of the intermediate model; If the processing accuracy is greater than or equal to an accuracy threshold, the intermediate model is determined as the target model; If the processing accuracy is less than the accuracy threshold, the intermediate model is updated according to the processing accuracy to obtain the target model.

5. The method according to any one of claims 1 to 3, characterized in that, The method further comprises the following steps: A weight value of each target parameter type in the target model is determined, and the weight value is used to indicate an influence degree on the task execution result; The weight value corresponding to each target parameter type is displayed.

6. A method of determining a result of a task execution, characterized by, The method comprises the following steps: A plurality of target parameter types corresponding to a target task are determined; A plurality of task parameters of the target task are obtained according to the plurality of target parameter types; The plurality of task parameters are processed by a target model to obtain a task execution result of the target task, and the target model is determined according to the method in any one of claims 1-5.

7. A model training apparatus, comprising: The method comprises the following steps: A first determining module, a first obtaining module, a second determining module, a second obtaining module and a training module are included, wherein The first determining module is used to determine a target task type, and determine a plurality of historical tasks according to the target task type, and the task type of the historical task is the target task type; the target task type includes a software development type and a product development type; The first obtaining module is used to obtain a task execution result of each historical task and a plurality of parameter values corresponding to a plurality of parameter types; if the historical task is a software development task, the task execution result includes success or failure; if the historical task is a product development task, the task execution result includes early completion, on-time completion and overdue completion; The second determining module is used to determine a plurality of target parameter types in the plurality of parameter types according to the task execution result of each historical task and the plurality of parameter values corresponding to the plurality of parameter types, and the correlation degree between the target parameter type and the task execution result is greater than or equal to a preset threshold; The second obtaining module is used to obtain a plurality of training samples according to the plurality of target parameter types, and the training sample includes a plurality of sample task parameters of a sample task and a sample execution result of the sample task, the task type of the sample task is the target task type, and the parameter type of the sample task parameter is the target parameter type; The training module is used to train a to-be-trained model according to the plurality of training samples to obtain a target model, and the target model is used to predict the task execution result of the target task. The second determining module is specifically configured to: perform classification processing on the plurality of historical tasks according to task execution results of each historical task, to obtain a plurality of historical task sets, the historical task sets being sets of historical tasks with the same task execution result; and for any one historical task set, determining, from the plurality of parameter types, a target parameter type corresponding to the historical task set according to a plurality of parameter values of the historical tasks in the historical task set. The plurality of target parameter types include the target parameter type corresponding to each historical task set.

8. An electronic device, comprising: Comprise: A processor, and a memory connected with the processor in communication; The memory stores computer-executable instructions; The processor executes the computer-executable instructions stored in the memory to implement the method of any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer-executable instructions, and the computer-executable instructions are executed by the processor to implement the method of any one of claims 1 to 6.

10. A computer program product comprising a computer program, characterized in that, The computer program is executed by the processor to implement the method of any one of claims 1 to 6.

Images