A Novel Approach in Clustering via Rough Set

Clustering is a widely used technique in data mining application for discovering patterns in large dataset. K-Means and Fuzzy C-Means clustering algorithm are the traditional approach for clustering. Above mentioned algorithm are analyzed and found two drawbacks that the quality of resultant cluster is based on prior fixation of cluster size K and on sequentially or randomly selected initial seed. Our earlier proposes namely UCAM (Unique Clustering with Affinity Measures) and Fuzzy-UCAM over bridged the drawbacks of K-Means and Fuzzy C-Means. UCAM and Fuzzy-UCAM clustering algorithm works without initial seed and prior fixation on number of clusters, where the unique clustering is obtained with the help of affinity measures. In this paper Rough Set Attribute Reduction (RSAR) is hybridized with UCAM and Fuzzy-UCAM, which reduces the computational complexity, increases the cluster Uniqueness, and retains the originality of the data.