Subspace clustering identifies the clusters stored in subspaces of a high dimensional dataset. Various Density-based strategies have been determined to mine clusters of arbitrary shape successfully even in the appearance of noise in full dimensional space clustering techniques. The performance and result of a subspace clustering algorithm highly depend on the parameter values of the algorithm is tuned to execute. Although determining the proper parameter values are crucial for both clustering

more »

... ality and performance of the algorithm. Further, as high dimensional data has converted more and more prevalent in real-world applications due to the advances of vast data technologies. Precisely Density-based subspace clustering have gained their importance owing to their ability to identify arbitrary shaped subspace clusters. Density Divergence Query is an essential subject in high dimensional data clustering. Density divergence involves having various subspace cardinalities for complex region densities. To defeat this problem, Efficient-EnSubClu employs an efficient subspace clustering model. It discovers the clusters using different epsilon density thresholds in various subspaces. In this research, we propose an Efficient enhanced Subspace Clustering Model named Efficient-EnSubClu (Enhancement of ENSUBCLU) for discovering precise values of parameters in subspace clustering. It allows efficient neighboring core points to be clustered and find quality subspace clusters satisfying specific qualitative and quantitative properties. Furthermore, apply the post-processing clustering steps on each found subspaces. It aims a merging model at the first step of evaluation of clusters connected with DBSCAN algorithm. Also, find the number of subspace clusters in a particular dimension and calculate the low mean dimensionality of subspace clusters. It represents every cluster with a fewer number of dimensions as visible from the low utility of mean dimensionality. Hence, we can obtain knowledge more concisely with an enhanced quality of clusters in terms like Accuracy and Silhouette Coefficient.