Open AccessA stable feature selection method based on relevancy and redundancy
Author(s) -
Peng Shen,
Xiaoming Ding,
Wenjun Ren,
Shu Liu
Publication year - 2021
Publication title -
journal of physics. conference series
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.21
H-Index - 85
eISSN - 1742-6596
pISSN - 1742-6588
DOI - 10.1088/1742-6596/1732/1/012023
Subject(s) - minimum redundancy feature selection , redundancy (engineering) , feature selection , computer science , data mining , software , uncorrelated , artificial intelligence , pattern recognition (psychology) , feature (linguistics) , data redundancy , selection (genetic algorithm) , stability (learning theory) , perspective (graphical) , machine learning , algorithm , mathematics , statistics , linguistics , philosophy , programming language , operating system
In this paper, the characteristics of software defect prediction are analyzed from the perspective of machine learning. To solve the problem of some redundant or uncorrelated features in defect data sets, a stable feature selection method based on relevancy and redundancy (RRSFS) is proposed. RRSFS combines the redundancy between features and the correlation between features and classes to select the optimal subset. RRSFS not only reduces the cost of data operation in the prediction model, but also enhances the stability of feature selection algorithm.