A granular computing paradigm for concept learning

Abstract

The problem of concept formation and learning is examined from the viewpoint of granular computing. Correspondences are drawn between granules and concepts, between granulations and classifications, and between relations over granules and relations over concepts. Two learning strategies are investigated. A global attribute-oriented strategy searches for a good partition of a universe of objects and a local attribute-value-oriented strategy searches for a good covering. The proposed granular computing paradigm for concept learning offers twofold benefits. Results from concept formulation and learning enrich granular computing and a granular computing viewpoint sheds new light on concept formulation and learning. © Springer-Verlag Berlin Heidelberg 2013.