Computational Theory Perception (CTP), Rough-Fuzzy Uncertainty Analysis and Mining in Bioinformatics and Web Intelligence: A Unified Framework

Abstract

Abstract. The concept of computational theory of perceptions (CTP), its characteristics and the relation with fuzzy-granulation (f-granulation) are explained. Role of f-granulation in machine and human intelligence and its modeling through rough-fuzzy integration are discussed. The Significance of rough-fuzzy synergestic integration is highlighted through three examples, namely, rough-fuzzy case generation, rough-fuzzy c-means and rough-fuzzy c-medoids along with the role of fuzzy granular computation. Their superiority, in terms of performance and computation time, is illustrated for the tasks of case generation (mining) in large-scale case-based reasoning systems, segmenting brain MR images, and analyzing protein sequences. Different quantitative measures for rough-fuzzy clustering are explained. The effectiveness of rough sets in constructing an ensemble classifier is also illustrated in a part of the article along with its performance for web service classification. The article includes some of the existing results published elsewhere under different topics related to rough sets and attempts to integrate them with CTP in a unified framework providing a new direction of research.