Entropy-Based Feature Selection for Data Clustering Using k-Means and k-Medoids Algorithms