Abstract

In data mining the Feature selection is the processof selecting a subset of relevant features for use in model construction. Thecentral assumption when using a feature selection technique is that the datacontain many redundant or irrelevant features. Of the many feature subsetselection algorithms, some can effectively eliminate irrelevant features butfail to handle redundant features yet some of others can eliminate theirrelevant while taking care of the redundant features. Fast clustering basedfeature selection algorithm is proposed. Features are different clusterrelatively independent. Clustering based strategy has high probability ofproducing a subset of important and independent features. To adopt theefficiency of fast clustering feature selection algorithm. It creates efficientminimum spanning tree clustering method.

Keywords-subset selection, clustering, Fast Algorithm