Granular computing in decision approximation. An application of rough mereology (Q2343488)

The book provides an extended presentation of granular computing, focusing on applications in classification/decision theory. The authors consider rough mereology as a good approach to granular computing, allowing for a good formalization of granules. The book drives readers from the introductory notions regarding similarity and granulations, passing through rough mereology, learning data classification, methodologies for granular reflections, covering strategies, layered granulation, etc., ending with effects of granulation on entropy and noise in data. Applications envisage experiments on 14 datasets from the UCI Data Repository. Overall, the book is intended to students and researchers interested in granular computing.