Fuzzy Clustering Applied on Mobile Agent Behaviour Selection
Eva Ocelíková · 2005
The goal od this paper is to provide a concept of possible hierarchical structructure of mechanizm for mobile agent behaviour selection. Hierarchy of the controler is composed of fuzzy classifier and several simple fuzzy systems, where each can deal with one simple situation. Fuzzy systems are able to treat and cope with uncertain and imprecise information, they have capability of expressing the knowledge in the form of linguistic rules, which sometimes are not easy to design. Fuzzy classifiers on the other hand do not need require antecendent knowledge of the space where they operate. Combination of those two techiques can leed into simple and effective system, where fuzzy classifier is employed to select the appropriate behavior, with wich one of the fuzzy controlers can deal with.