K-Means vs. Fuzzy C-Means: A Comparative Analysis of Two Popular Clustering Techniques on the Featured Mobile Applications Benchmark

Abstract

Over the past few years, mobile applications have become an indispensable part of our daily lives. Noticing this ever- growing market, all those who are engaged in developing attractive applications should make informed decisions along the development process through sophisticated methods in order to survive in the market. As one of these methods, clustering is well suited for identifying the hidden groups existing in huge datasets. In this paper, the Mobile App dataset that contains features of 7196 available applications was clustered using two popular clustering algorithms, namely as k-means and fuzzy-c means. After conducting necessary preprocessing steps (e.g. outlier removal, standardization), these algorithms were run with different parameters in an experimental manner to reach optimal values and their performances were compared based on cluster quality (internal validity), number of iterations and elapsed time. The main findings suggested that fuzzy c- means produced higher quality clusters whereas k-means algorithm converged faster than its counterpart. In the last section, conclusions were made and future studies were discussed.


Editor: H. Kemal İlter, Ankara Yıldırım Beyazıt University, Turkey

Received: August 19, 2018, Accepted: October 18, 2018, Published: November 10, 2018
Copyright: © 2018 IMISC Hakyemez et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.