Clustering Medical Data Using Subspace and Parallel Approximation Algorithm

B. Thenmozhi, P. Shanthi

Abstract: In high-dimensional feature spaces traditional clustering algorithms tend to break down in terms of efficiency and quality. Nevertheless, the data sets often contain clusters which are hidden in various subspaces of the original feature space. In high dimensional data, however, many of the dimensions are often irrelevant. These irrelevant dimensions confuse clustering algorithms by hiding clusters in noisy data. In this paper we propose parallel approximation algorithm localize the search for relevant dimensions allowing them to find clusters that exist in multiple, possibly overlapping subspaces. A broad evaluation based on real-world medical data sets demonstrates that is suitable to find all relevant subspaces in high dimensional, sparse data sets and produces better results than existing methods.

Keywords: Subspace clustering, Dimensionality Reduction, Redundancy Awareness, Detecting Relevant Attributes, Greedy optimization