Genetic K-Means Algorithm – Implementation and Analysis
Sonia Sharma1, ShikhaRai2
1Mrs Sonia Sharma, Department of Computer Science JMIT, Radaur, Yammuna Nagar, Haryana, India.
2Miss Shikha Rai, Department of Computer Science JMIT, Radaur, Yammuna Nagar, Haryana, India.
Manuscript received on 18 June 2012 | Revised Manuscript received on 25 June 2012 | Manuscript published on 30 June 2012 | PP: 117-120 | Volume-1 Issue-2, June 2012 | Retrieval Number: B0227051212/2012©BEIESP
Open Access | Ethics and Policies | Cite | Mendeley | Indexing and Abstracting
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC-BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Abstract: K-means algorithm is most widely used algorithm for unsupervised clustering problem. Though it is accepted but it has some problems which make it unreliable. Initialization of the random cluster centres, number of clusters and terminating condition play a major role in quality of clustering achieved. This paper empirically analyses a derived form [Krishna &Narasimha, 1999] of K-means using Genetic algorithm approach. The new algorithm prevents algorithm to converge towards local minima by considering a rich population of potential solutions. A tool that implements this algorithm is presented in the paper. The time complexity and execution expectation is also tested over an exhaustive set of data of different dimensions.
Keywords: K-Means Clustering, Genetic Algorithm, Local Minima, Optimization.
Scope of the Article: Optimization