Efficient Clustering of high dimensional nonlinear data using modified Denclue algorithm

Abstract

Clustering is a data mining task devoted to the automatic grouping of data, based on mutual similarity. Clustering in high-dimensional spaces is a recurrent problem in many domains. It affects time complexity, space complexity, Data Size Adaptability and Precision Value of clustering methods. High-dimensional data usually live in different low dimensional subspaces hidden in the original space. As high‐dimensional objects appear almost alike, new approaches for clustering are required. Consequently, recent research has focused on developing techniques and clustering algorithms specifically for high‐dimensional data.

In this research work, Denclue is chosen to analyse the compatibility of clustering high dimensional data. Denclue is a density based clustering technique. In density based clustering, the objects are classified based on their regions of density. These algorithms have the ability to discover classes of arbitrary shapes and omit noisy objects. To efficiently cluster the high dimensional nonlinear data, a new modified Denclue is proposed by incorporating the mathematics methods such as meta-heuristics, curse of dimensionality, sub spaces, data routing, correlation, normal distribution and darboux variate. The advantages of this proposed algorithm are it works on erroneous data, noisy data, provides better Clustering Pace, Competence Rate, Data Size Adaptability, Precision Value and Prognostic Reliability.