Application of Harmony Search Algorithm on Clustering
Babak Amiri, Liaquat Hossain, Seyyed Esmaeil Mosavi · 2010
Abstract — The harmony search is considered as musician's behavior which is inspired by soft computing algorithm. As the musicians in improvisation process try to find the best harmony in terms of aesthetics, the decision variables in optimization process try to be the best vector in terms of objective function. Cluster analysis is one of attractive data mining technique that use in many fields. One popular class of data clustering algorithms is the center based clustering algorithm. K-means used as a popular clustering method due to its simplicity and high speed in clustering large datasets. However, k-means has two shortcomings: dependency on the initial state and convergence to local optima and global solutions of large problems cannot found with reasonable amount of computation effort. In order to overcome local optima problem lots of studies done in clustering. This paper describes a new clustering method based on the harmony search (HS) meta-heuristic algorithm, which was conceptualized using the musical process of searching for a perfect state of harmony. The HS algorithm does not require initial values and uses a random search instead of a gradient search, so derivative information is unnecessary. We compared proposed algorithm with other heuristics algorithms in clustering, such as GA, SA, TS, and ACO, by implementing them on several simulation and real datasets. The results indicate that the proposed clustering is a powerful clustering method suggesting higher degree of precision and robustness than the existing algorithms. Index Terms — clustering, meta-heuristic, k-means, harmony search algorithm.