2016 Information Technology and Mechatronics Engineering Conference (ITOEC 2016)

45 / 2015-12-30 08:33:04

Improved Mutual Information Based on Relative Frequency Factor and Degree of Difference among Classes

MI,feature,selection,relative,frequency,factor,degree,of,difference,among,classes

By introducing degree of difference among classes, an improved mutual information feature selection method is proposed to effectively improve the accuracy of feature selection, accuracy and efficiency of classification. At the same time, relative frequency factor is applied to solve the tendency of traditional methods to choose the shortage of low frequency words. The experimental results show that the improved method can reasonably improve the performance of mutual information feature selection.