Rapid Reduction Algorithm Based on the Conditional Information Quantity
Yushu Liu
Abstract
Yushu Liu
Abstract
To improve the efficiency of attribute reduction,a rapid reduction algorithm based on conditional information quantity is proposed.The concepts of information quantity based on partition and conditional information quantity based on partition are defined,and the theorems about the monotone of attribute significance、invariance of conditional information quantity based on partition are proven;Compared with other algorithms,this algorithm reduces the searching space of attributes and samples in each step.Experimental results showed that the rapid reduction algorithm is more efficient than the existing algorithms.
OpenAlex reports 2 citations for this work. Citation counts describe recorded attention and do not establish research quality.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
To improve the efficiency of attribute reduction,a rapid reduction algorithm based on conditional information quantity is proposed.The concepts of information quantity based on partition and conditional information quantity based on partition are defined,and the theorems about the monotone of attribute significance、invariance of conditional information quantity based on partition are proven;Compared with other algorithms,this algorithm reduces the searching space of attributes and samples in each step.Experimental results showed that the rapid reduction algorithm is more efficient than the existing algorithms.
Key concepts: Partition (number theory), Reduction (mathematics), Algorithm, Monotone polygon, Mathematics, Conditional mutual information, Partition problem, Computer science