Title :
Importance Degree of Features and Feature Selection
Author :
Xiao, Di ; Zhang, Junfeng
Author_Institution :
Sch. of Autom. & Electr. Eng., Nanjing Univ. of Technol., Nanjing, China
Abstract :
A novel measure, importance degree of features, is proposed to rank the features. And a new filter method is presented to carry out feature selection based on such measure. The monotonic property of this proposed measure can reduce the search space, which results in enhancing learning efficiency. The simulation results indicate the validity of our method.
Keywords :
pattern recognition; unsupervised learning; feature filter method; feature importance degree; feature selection; machine learning; monotonic property; search space reduction; Automation; Computational efficiency; Computational modeling; Educational institutions; Electric variables measurement; Extraterrestrial measurements; Filters; Fuzzy systems; Space technology; Support vector machines; Feature Ranking; Feature Selection; Importance Degree of Features;
Conference_Titel :
Fuzzy Systems and Knowledge Discovery, 2009. FSKD '09. Sixth International Conference on
Conference_Location :
Tianjin
Print_ISBN :
978-0-7695-3735-1
DOI :
10.1109/FSKD.2009.625
