Unlock instant, AI-driven research and patent intelligence for your innovation.

Parameter optimization method for classification model, device, computer equipment and storage medium

A technology of classification model and optimization method, which is applied in computer parts, calculation, character and pattern recognition, etc., can solve problems such as the inability to improve the usability of the classification model, increase the time complexity of the classification model, and high threshold

Inactive Publication Date: 2018-01-19
SHENZHEN LEXIN SOFTWARE TECH CO LTD
View PDF0 Cites 10 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] The Xgboost classification model is a commonly used model in the prediction method, but the Xgboost classification model requires the user to manually set and adjust the construction parameters required for the model in the actual use process, thus setting a high threshold for the user; in addition , based on the existing parameter optimization methods, such as grid search, when optimizing the construction parameters of the Xgboost classification model, the optimization method is often based on the empirically given or exhaustive search space, which will not only fail to improve the performance of the classification model. Ease of use increases the time complexity of the classification model

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Parameter optimization method for classification model, device, computer equipment and storage medium
  • Parameter optimization method for classification model, device, computer equipment and storage medium
  • Parameter optimization method for classification model, device, computer equipment and storage medium

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0028] figure 1 It is a schematic flowchart of a parameter optimization method for a classification model provided in Embodiment 1 of the present invention. This method is suitable for optimizing the construction parameters required for the construction of the classification model. The method can be executed by a parameter optimization device for the classification model, wherein The device can be implemented by software and / or hardware, and is generally integrated on computer equipment.

[0029] It should be noted that the application background of this embodiment may be: using a classification model to predict the churn of consumer users in the e-commerce platform. Generally, the classification model usually needs to be constructed with given construction parameters, and then the constructed classification model can be trained and learned with given training samples, and finally a practically applicable classification model can be obtained.

[0030] Therefore, before using ...

Embodiment 2

[0046] figure 2 It is a schematic flowchart of a parameter optimization method for a classification model provided in Embodiment 2 of the present invention. Embodiment 2 of the present invention is optimized on the basis of the above-mentioned embodiments. In this embodiment, the component values ​​in the parameter correlation vectors and the moving speeds corresponding to the component values ​​are further initialized, specifically It is: randomly select a numerical value within the set first value range as the initial component value of each dimension in each parameter-related vector; randomly select a numerical value within the set second value range as the initial value of each said component value corresponding movement speed.

[0047] At the same time, each of the initial parameter related vectors will be updated iteratively according to the set update strategy to obtain a globally optimal target parameter related vector, which is further embodied as: using each of the...

Embodiment 3

[0103] image 3 A structural block diagram of a device for optimizing parameters of a classification model provided in Embodiment 3 of the present invention. The device is suitable for optimizing the construction parameters required for the construction of the classification model, the device can be realized by software and / or hardware, and is generally integrated on computer equipment. Such as image 3 As shown, the device includes: a parameter vector construction module 31 , a parameter vector initial module 32 , a target vector determination module 33 and an optimal parameter determination module 34 .

[0104] Wherein, the parameter vector construction module 31 is used to determine the number of parameters of the construction parameters in the classification model to be constructed, and generates a parameter correlation vector whose dimension of the set number is the number of parameters;

[0105] A parameter vector initialization module 32, configured to initialize each...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

The invention discloses a parameter optimization method for a classification model, a device, computer equipment and a storage medium. The method comprises the following steps of determining the number of constructed parameters in a to-be-constructed classification model, generating parameter-related vectors with the dimensions of the preset number thereof to be equal to the number of parameters;initializing each parameter-related vector to obtain a preset number of initial parameter-related vectors containing the initial component information; according to a preset updating strategy, iteratively updating each initial parameter-related vector to obtain a target parameter correlation vector containing the global optimum component information; according to the global optimum component information, determining an optimal parameter value of each constructed parameter. Based on the method, the optimal parameter value of each constructed parameter required for the consecution of the classification model can be automatically determined. Therefore, the classification model with higher working performance can be constructed on the premise that the time complexity of the classification model is not increased. The use threshold of the classification model is reduced, and the user experience of the prediction of the classification model is improved.

Description

technical field [0001] The invention relates to the technical field of classification prediction, in particular to a parameter optimization method, device, computer equipment and storage medium of a classification model. Background technique [0002] Generally speaking, any product of any company will lose users, which is inevitable in the replacement of old and new users. Due to the existence of customer acquisition costs, compared with maintaining the activity of old users, it is bound to cost more to activate the lost users immediately. Therefore, it is very necessary to predict the possibility of future loss of existing users through technical means in advance, and to implement corresponding marketing means to retain them in time before the loss stage of the user life cycle. [0003] Forecasting the loss of e-commerce platform users is equivalent to analyzing the users who have already consumed on the e-commerce platform to determine whether these users may consume agai...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
Patent Type & Authority Applications(China)
IPC IPC(8): G06K9/62G06Q30/02
Inventor 吴佳东
Owner SHENZHEN LEXIN SOFTWARE TECH CO LTD