Gene expression data analyses for supervised prostate cancer classification based on feature subset selection combined with different classifiers

Abstract

In machine learning, feature selection is the process of selecting a subset of relevant features for use in model construction. A comparative evaluation between selection methods: SNR, Correlation Coefficient and Max-relevance Min-Redundancy is carried out, using the dataset of prostate cancer. The Evaluation of the dimensionality reduction was done by using the supervised classifier K Nearest Neighbors (KNN), Support Vector Machines (SVM), Linear Discriminant Analysis (LDA) and Decision Tree for supervised classification (DTC). The purpose of classification is to assign an object to a certain class. The classifier shows that the combination between SNR and the LDA classifier can present the highest accuracy.