Abstract: Clustering large data set of high dimensionality has always been a serious challenge for clustering algorithms. Traditional clustering algorithms often fail to detect meaningful clusters because of the high dimensionality and inherently sparse feature space of most real-world data sets.Nevertheless, the data sets often contain clusters hidden in various subspaces of the original feature space.In addition, high-dimensional data often contain a significant amount of noise which causes additional effectiveness problems.To overcome these problems, a new algorithm based on CLIQUE, named OpCluster, is proposed.A set of experiments on a synthetic dataset demonstrate the effectiveness and efficiency of the new approach.

Key words: clustering algorithms, subspace clustering, optimal partition, data partition