Fuzzy sets in pattern recognition: methodology and methods

摘要: Abstract An objective of the paper is to discuss a state-of-the-art methodology and algorithms fuzzy sets in field pattern recognition. In real-world recognition classification problems we are faced with fuzziness that connected diverse facets cognitive activity human being. origin sources related labels expressed feature space as well classes taken into account procedures. evident difference between way information processing by means probability set theory interpretation results explained detail. sequel methods studied two main streams, namely supervised unsupervised learning. Different approaches designing schemes (as e.g. relation calculus, decision-making approach, etc.) put account. A method selection aid measure integral introduced. clustering techniques relying on minimization function analyzed Also problem cluster validity terms indices addressed. Numerical examples also provided.