An efficient & effective feature subset selection for high dimensional data

Abstract

Feature selection process involves identifying a subset of the most useful features that produces compatible results as the original entire set of features. The feature subset selection reduces the data size by removing irrelevant features and eliminating redundant features and improves the classifier accuracy. The feature selection from high dimensional data is challenging task due significant large number of features. The feature selection algorithm is constructed with the consideration of efficiency(Time) and effectiveness(Accuracy) point of view. The proposed feature selection process carried out using graph theoretic method i.e. FAST algorithm. It provides novel way for irrelevant feature removal followed by redundant feature elimination with the help of Kruskal's minimum spanning tree clustering method.