अमूर्त

Exploring Constraints Inconsistence for Value Decomposition and Dimension Selection Using Subspace Clustering

K.Prema, K.Sangeetha, Dr.S.Karthik

The datasets which are in the form of object-attribute-time is referred to as threedimensional (3D) data sets. As there are many timestamps in 3D datasets, it is very difficult to cluster. So a subspace clustering method is applied to cluster 3D data sets. Existing algorithms are inadequate to solve this clustering problem. Most of them are not actionable (ability to suggest profitable or beneficial action), and its 3D structure complicates clustering process. To cluster these three-dimensional (3D) data sets a new centroid based concept is introduced in the proposed system called PCA. This PCA framework is introduced to provide excellent performance on financial and stock domain datasets through the unique combination of Singular Value Decomposition, Principle Component Analysis and 3D frequent item set mining.PCA framework prunes the entire search space to identify the significant subspaces and clusters the datasets based on optimal centroid value. This framework acts as the parallelization technique to tackle the space and time complexities.