Time Complexity of Rough Clustering: GAs versus K-Means

Abstract

Typical clustering operations in data mining involve finding natural groupings of resources or users. Conventional clusters have crisp boundaries, i.e. each object belongs to only one cluster. The clusters and associations in data mining do not necessarily have crisp boundaries. An object may belong to more than one cluster. Researchers have studied the possibility of using fuzzy sets in data mining clustering applications. Recently, two different methodologies based on properties of rough sets were proposed for developing interval representations of clusters. One approach is based on Genetic Algorithms, and the other is an adaptation of K-means algorithm. Both the approaches have been successful in generating intervals of clusters. The efficiency of the clustering algorithm is an important issue when dealing with a large dataset. This paper provides comparison of the time complexity of the two rough clustering algorithms.