Title :
Optimal feature selection based on Bayesian networks
Author :
Zhao, Hui ; Ming Xiao ; Xiao, Ming
Author_Institution :
Central China Normal Univ., Wuhan
Abstract :
Focusing on feature selection in classification, we investigate some popular methods which are based on learning the structure of Markov blanket for the classification variable in a Bayesian network and propose two new algorithms. Initial experiments demonstrate their advantages over some existing well known methods.
Keywords :
Markov processes; belief networks; pattern classification; Bayesian networks; Markov blanket; optimal feature selection; pattern classification; Algorithm design and analysis; Bayesian methods; Information analysis; Input variables; Iterative algorithms; Mathematics; Notice of Violation; Pattern analysis; Pattern recognition; Wavelet analysis; Bayesian networks; Feature selection; Markov blanket; conditional independence;
Conference_Titel :
Wavelet Analysis and Pattern Recognition, 2007. ICWAPR '07. International Conference on
Conference_Location :
Beijing
Print_ISBN :
978-1-4244-1065-1
Electronic_ISBN :
978-1-4244-1066-8
DOI :
10.1109/ICWAPR.2007.4420739
