Tumor Classification by Using PCA with Hybrid Feature Selection Mechanism

Feature extraction is an important issue for analysis of gene expression microarray data, of which principle component analysis (PCA) is one of the frequently used methods, and in the previous works, the top several principle components are selected for modeling according to the descending order of eigenvalues. In this paper, we argue that not all the first features are useful, but features should be selected form all the components by feature selection methods. We demonstrate a framework for selecting good feature subsets from all the principle components, leading to reduced classifier error rates on the microarray data. As a case study, we have considered PCA for feature extraction; then the Hybrid Feature Selection Mechanism filter model are used to select the top features extracted by PCA, in the first stage, we rank the features by the Relief algorithm between each feature and each class, and then choose the highest relevant features to the classes with the help of the threshold, in the second stage, we use Shepley value to evaluate the contribution of features to the classification task in the ranked feature subset; At last, SVMs are used as classifiers. Experimental results illustrate that our proposed framework is effective to reduce classification error rates. (C) 2011 Published by Elsevier Ltd. Selection and/or peer-review under responsibility of the Organizers of 2011 International Conference on Energy and Environmental Science.