TY - JOUR
T1 - A hybrid feature selection algorithm for gene expression data classification
AU - Lu, Huijuan
AU - Chen, Junying
AU - Yan, Ke
AU - Jin, Qun
AU - Xue, Yu
AU - Gao, Zhigang
N1 - Funding Information:
This study is supported by National Natural Science Foundation of China (No. 61272315, No. 61303183, No.61402417, No. 61602431, No. 60842009, and No. 60905034), Zhejiang Provincial Natural Science Foundation (No. Y1110342, No. LY15F020037, and No. Y1080950) and Zhejiang Provincial Science and Technology Department of International Cooperation Project (No. 2012C24030). It is also supported by the Priority Academic Program Development of Jiangsu Higher Education Institutions (PAPD) and Jiangsu Collaborative Innovation Center on Atmospheric Environment and Equipment Technology (CICAEET).
Publisher Copyright:
© 2017
PY - 2017/9/20
Y1 - 2017/9/20
N2 - In the DNA microarray research field, the increasing sample size and feature dimension of the gene expression data prompt the development of an efficient and robust feature selection algorithm for gene expression data classification. In this study, we propose a hybrid feature selection algorithm that combines the mutual information maximization (MIM) and the adaptive genetic algorithm (AGA). Experimental results show that the proposing MIMAGA-Selection method significantly reduces the dimension of gene expression data and removes the redundancies for classification. The reduced gene expression dataset provides highest classification accuracy compared to conventional feature selection algorithms. We also apply four different classifiers to the reduced dataset to demonstrate the robustness of the proposed MIMAGA-Selection algorithm.
AB - In the DNA microarray research field, the increasing sample size and feature dimension of the gene expression data prompt the development of an efficient and robust feature selection algorithm for gene expression data classification. In this study, we propose a hybrid feature selection algorithm that combines the mutual information maximization (MIM) and the adaptive genetic algorithm (AGA). Experimental results show that the proposing MIMAGA-Selection method significantly reduces the dimension of gene expression data and removes the redundancies for classification. The reduced gene expression dataset provides highest classification accuracy compared to conventional feature selection algorithms. We also apply four different classifiers to the reduced dataset to demonstrate the robustness of the proposed MIMAGA-Selection algorithm.
KW - Adaptive genetic algorithm
KW - Feature selection
KW - Gene expression data
KW - Mutual information maximization
UR - http://www.scopus.com/inward/record.url?scp=85014772083&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85014772083&partnerID=8YFLogxK
U2 - 10.1016/j.neucom.2016.07.080
DO - 10.1016/j.neucom.2016.07.080
M3 - Article
AN - SCOPUS:85014772083
SN - 0925-2312
VL - 256
SP - 56
EP - 62
JO - Neurocomputing
JF - Neurocomputing
ER -