Score-Based Feature Selection of Gene expression Data for Cancer Classification

Abstract

Feature selection in machine learning can also be specified as attribute selection. It is a process of selection desired feature from a large amount of data set. A typical microarray data set has basic properties such as high-dimensionality and limited sample, which makes it less accurate for classification and also time-consuming. In order to increase the accuracy of the classification, we have to decrease the dimensionality of the dataset. To achieve this, there are two feature elimination methods namely, feature selection and feature extraction. The proposed study focuses on the filter-based feature selection method. The main aim of the proposed work is to decrease the computation time and increase the accuracy of classification and prediction. To achieve this, he proposed work reduces the dimensionality of data set and also the redundancy between various features. Several feature selection methods exist but most of them have increased computational time, so here we are using score-based criteria fusion method for feature selection, which improves the prediction accuracy and decreases the computational time.