Model feature analysis method and device, equipment and medium
A feature analysis and model technology, applied in the computer field, can solve the problems of large model size and high calculation pressure, and achieve the effects of controlling model size, reducing online storage, and improving accuracy
Active Publication Date: 2020-03-24
BIGO TECH PTE LTD
View PDF1 Cites 3 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
[0005] In view of this, the embodiment of the present invention provides a model feature analysis method, device, equipment and medium, which uses the sample set in the log as the calculation object for analysis to ensure The analysis is based on the data distribution of the actual prediction set, which improves the accuracy of the analysis results, thereby solving the problems of online storage and calculation pressure caused by the large model size in the prior art
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 moreImage
Smart Image Click on the blue labels to locate them in the text.
Smart ImageViewing Examples
Examples
Experimental program
Comparison scheme
Effect test
Embodiment Construction
[0023] The present invention will be further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described here are only used to explain the present invention, but not to limit the present invention. In addition, it should be noted that, for the convenience of description, only parts related to the present invention are shown in the drawings but not all structures or components.
[0024] refer to figure 1 , which shows a schematic flow chart of the steps of an embodiment of a model feature analysis method of the present invention, which may specifically include the following steps:
[0025] Step 110, acquire the sample information in the log, which is generated when the target model is making online prediction.
[0026] Specifically, each time the target model is tested online, a corresponding log can be generated. The log can be used as a sample set, and can specifically contain one ...
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
Login to View More Abstract
The invention discloses a model feature analysis method and device, equipment and a medium, and relates to the technical field of computers. The model feature analysis method comprises the steps: acquiring sample information in a log, wherein the log is generated during online prediction of a target model; disturbing target features in the sample information to obtain feature disturbance information; determining sorting change information corresponding to the sample information according to the characteristic disturbance information; and determining an importance analysis result of the targetfeature according to the sorting change information. According to the model feature analysis method, the accuracy of an analysis result is improved, and feature screening can be effectively helped, and the model scale can be effectively controlled, so that the pressure of online storage and calculation is effectively reduced.
Description
technical field [0001] The present invention relates to the field of computer technology, in particular to a model feature analysis method, device, equipment and medium. Background technique [0002] With the rapid development of computer technology, the depth sorting model is becoming more and more popular. Compared with models such as logistic regression and decision trees, the deep sorting model is more "black box", less analytical, and more difficult to analyze the importance of different features in the deep sorting model, so it is difficult to control deep sorting by screening features The scale of the model leads to online storage of the model and heavy calculation pressure. How to analyze the importance of each feature in the deep sorting model has become a hot spot for many researchers. [0003] The existing methods for analyzing the importance of the depth sorting model usually adopt the following two methods: one is the Ablation Test method, which removes one fe...
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
Login to View More Patent Type & Authority Applications(China)
IPC IPC(8): G06F16/22
CPCG06F16/2228
Inventor 陈华峰
Owner BIGO TECH PTE LTD