Covering Rough Clustering Approach for Unstructured Activity Analysis

Several tasks under human activities need to be performed in a sequence of navigation and manipulation of objects. In several applications of human activities like robotics monitoring plays an important role. So, in these applications, processing of sequential data is of utmost importance. Because of the presence of imprecision intelligent clustering approaches using fuzzy or rough set techniques play a major role. The basic rough sets which are defined by using equivalence relations is less useful because of their scarcity in real life scenarios. As a result, covering based rough sets have been introduced which are more general and applicable to real world problems. In this paper, covering rough set based clustering approach is introduced and studied using refined first type of covering based rough sets. Through experimental analysis,illustrated the efficiency of proposed algorithm and provided a comparative analysis of this algorithm with other existing algorithms.