Projective Clustering Approach For The Detection Of Outlier And Non-Axis-Aligned Subspaces

Abstract: Clustering the case of non-axis-aligned subspaces and detection of outliers is a major challenge due to the curse of dimensionality. To solve this problem, the proposed implementation is extension to traditional clustering and finds subsets of the dimensions of a data space .In this project, a probability model is proposed to describe in hidden views and the detection of possible selection of relevant views. A projective clustering is proposed for Outlier Detection in High Dimensional Dataset that discovers the detection of possible outliers and non-axis-aligned subspaces in a data set and to build a robust initial condition for the clustering algorithm it improves the parameters in the connection between L∞ corsets and sensitivity that is made in Lemma and improve clustering in the case of non-axis-aligned subspaces and detection of outliers in datasets. The suitability of the proposal demonstrated is done with synthetic data set and some widely used real-world data set.

Keywords: Clustering, high dimensions, projective clustering, probability model.