Ashish Parmar, Yugal Kumar, Pradeep Kumar Singh and Vijendra Singh
Pertanika Journal of Science & Technology, Volume 27, Issue 2, April 2019
Keywords: Clustering, K-means, meta-heuristic algorithm, vibrating particle system
Published on: 24 Apr 2019
In the field of data analysis, clustering is an unsupervised technique that can be used to find identical sets of data. But, it is tough task to find the optimal centroid for a given dataset, especially in hard clustering problems. Recently, a vibrating particle system (VPS) algorithm was developed for solving the optimization problems. This algorithm is based on the concept of free vibration and forced vibration. This algorithm provides more effective and optimal solutions for constrained optimization problems. In this work, the performance of VPS algorithm is evaluated for solving hard clustering problems. The objective of this algorithm is to compute optimal centroid for hard clustering problems. The efficiency of the proposed algorithm is measured on well known clustering datasets and compared with some popular clustering algorithms. The simulation results demonstrate that the VPS algorithm obtains effective results as compared to other algorithms.
Share this article