Data Classification Using Feature Selection and kNN Machine Learning Approach

Abstract

The k Nearest Neighbour (kNN) method is one of the most popular algorithm in clustering and data classification. The kNN algorithm founds to be performed very efficient in the experiments on different dataset. In this paper, we focus on the classification problem. The algorithm is experienced over Leukemia dataset. Initially three feature selection algorithm Consistency Based Feature Selection (CBFS), Fuzzy Preference Based Rough Set (FPRS) and Kernelized Fuzzy Rough Set (KFRS) is applied on the dataset and then kNN is applied as a classifier onto the dataset. The results of our experiment demonstrates that CBFS algorithm generally perform better than other two KFRS and FPRS algorithm respectively.