SVM-RF-based decision rule extraction and reduction method

A decision-making and rule-based technology, applied in instruments, character and pattern recognition, computer parts, etc., to achieve the effect of extracting and reducing the number of rules, fewer rules, and taking into account the accuracy rate

Inactive Publication Date: 2019-07-05
BEIJING INSTITUTE OF TECHNOLOGYGY
View PDF0 Cites 6 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Also some trees differ only in topo

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
  • SVM-RF-based decision rule extraction and reduction method
  • SVM-RF-based decision rule extraction and reduction method
  • SVM-RF-based decision rule extraction and reduction method

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0030] In order to better illustrate the purpose and advantages of the present invention, the implementation of the method of the present invention will be further described in detail below in conjunction with the accompanying drawings and examples.

[0031] The specific process is:

[0032] Assuming that a classification problem divides data into three categories (A, B, C), it is necessary to mine the hidden rules in the SVM-RF model, extract and reduce them in a human-understandable range. Record the data set as D 0 , whose sample size is n 0 , the number of features is k, and the sample form is i∈(0,n 0 ), where x i is the eigenvector of the i-th data, that is, x i =(x i1 ,x i2 … x ij ...x ik ), j∈[1,k], x ij is the j-th feature of the i-th data. the y i Is the label of the i-th data, which is one of the three categories (A, B, C).

[0033] Step 1, use the data set D to train the SVM model, and extract the data subset that can represent the decision boundary f...

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 an SVM-RF-based decision rule extraction and reduction method, and belongs to the technical field of computer and information science. The method comprises the following stepsof training an SVM by using data to obtain a classifier and a support vector; generating new data characteristics by adopting a regeneration tree method, obtaining a new data label by using an SVM (support vector machine), and integrating the new data to obtain a most information amount of data set; training a random forest model by using the data set with the maximum information amount to obtaina plurality of decision-making trees; fusing the terminal node similarity and the decision tree performance similarity of the decision tree into new similarity by introducing a trade-off factor, and reducing the redundant decision tree based on the similarity; and finally obtaining a rule set by using a decision tree traversal method. The decision rule extraction and reduction method provided by the invention not only gives consideration to higher accuracy of the SVM-RF model, but also can avoid that the extracted decisions are too many and are not easy to understand by people, thereby helpingthe SVM-RF model to popularize in practical application and playing a role in assisting human decision making.

Description

technical field [0001] The invention relates to a method for extracting and reducing decision rules based on SVM-RF, belonging to the fields of computer science and information technology. Background technique [0002] Rule extraction is to supplement the excellent performance of the black box model with an understandable rule set, expressing the hidden logical rules in the model in a way that is easy for humans to understand. In current scientific research, many decision-making classification models based on machine learning methods are called black boxes, that is, systems that hide their internal logic from users. Although this type of model has high performance, it cannot be used in real life because it will cause moral problems and unclear accountability mechanisms when using a decision-making model with an unclear internal logic mechanism in real life. Therefore, it has become an urgent problem to realize the rule extraction of such models to enhance the interpretabili...

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
IPC IPC(8): G06K9/62
CPCG06F18/2411G06F18/24323
Inventor 潘丽敏秦枭喃罗森林王海州
Owner BEIJING INSTITUTE OF TECHNOLOGYGY
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap