Clustering Efficiency of ACO and K-Harmonic Means Techniques: A Comparative Study

Abstract

In the last two decades, many advances on the computer sciences have been based on the observation and emulation of processes of the natural world. The nature inspired methods like ant based clustering techniques have found success in solving clustering problems. They have received special attention from the research community over the recent years because these methods are particularly suitable to perform exploratory data analysis. The clustering is an important technique that has been studied in various fields with many applications such as image processing, marketing, data mining and information retrieval. Recently, the various algorithms inspired by nature are used for clustering. Data clustering is a useful process to extract meaning from sets of unlabeled data or to perform data exploration for pattern recognition. This paper focuses on the behavior of clustering procedures in two approaches, ant based clustering algorithm and K-harmonic means clustering algorithm. The two algorithms were evaluated in two of well- known benchmark data sets. Empirical results clearly show that ant clustering algorithm performs well compared to another technique called K-Harmonic means clustering algorithm.