A Novel Two-Stage Cancer Classification Method for Microarray Data Based on Supervised Manifold Learning

Gene expression data analysis is a very useful tool for medical diagnosis. Combined with classification methods, this technology can be used to help make clinical decisions for individual patients. In this paper, a novel classification method for cancer microarray data was proposed. This method includes two stages: The first stage is to select a number of genes based on a gene selection algorithm, and then supervised locality preserving projections (SLPP) is accepted for further dimension reduction and discriminant feature extraction. This stage can find more discriminant projection direction based on training data. The second stage uses nearest neighborhood (NN) and support vector machine (SVM) for classification. To show the validity of the proposed method, 4 real cancer data sets were used for classifying. The prediction performance was evaluated by 3-fold cross validation. The experimental results show that the method presented here is effective and efficient.