A WK-means Approach for Clustering
Fatemeh Boobord1, Zalinda Othman2, and Azuraliza Abu Bakar3
1, 2,3Data Mining and Optimization Research Group, Center for Artificial Intelligence Technology, University Kebangsaan Malaysia, Malaysia
Abstract: Clustering is an unsupervised learning method that is used to group similar objects. One of the most popular and efficient clustering methods is K-means, as it has linear time complexity and is simple to implement. However, it suffers from gets trapped in local optima. Therefore, many methods have been produced by hybridizing K-means and other methods. In this paper, we propose a hybrid method that hybridizes Invasive Weed Optimization and K-means. The Invasive Weed Optimization algorithm is a recent population based method to iteratively improve the given population of a solution. In this study, the algorithm is used in the initial stage to generate a good quality solution for the second stage. The solutions generated by the Invasive Weed Optimization Algorithm are used as initial solutions for the K-means algorithm. The proposed hybrid method is evaluated over several real world instances and the results are compared with well-known clustering methods in the literature. Results show that the proposed method is promising compared to other methods.
Keywords: Data clustering, K-means algorithm, Invasive Weed Optimization, Hybrid evolutionary optimization algorithm, unsupervised learning
Received November 28, 2012; accepted August 12, 2013