A Pragmatic Application of Clustering-Based Feature Subset Selection Algorithm for High Dimensional Data

Using the rapid growth of computational biology and e-commerce applications, high-dimensional data becomes usual. Thus, mining high dimensional data is an urgent problem of great practical importance. Within the high dimensional data the dimensional reduction is a vital factor, to the purpose the clustering based feature subset selection algorithm is proposed in this particular paper. The characteristics are actually clustered Based on the class labels. The Relevance on the clustered features has become evaluated. The correlation on the relevant clustered feature will be evaluated. This technique improved by cluster based FAST Algorithm and Fuzzy Logic. FAST Algorithm can often Identify and taking out the irrelevant data set. This algorithm process implements using two different steps which are graph theoretic clustering methods and representative feature cluster is selected. Feature subset selection researchers have centered on in search of relevant features. The proposed fuzzy logic has focused on minimized redundant data set and improves the feature subset accuracy.