Bilkent ACM SIGART (Special Interest Group on Artificial Intelligence)

In this talk, after a brief introduction to fuzzy logic, approaches for adding a learning component to fuzzy systems to improve performance will be discussed. In the conducted literature survey, it was observed that mainly rote learning and inductive learning methods were used. Among these, supervised learning methods are more powerful. The prominent learning methods are hybrid supervised learning methods based on neuro-fuzzy architectures and genetic algorithms and fuzzy logic controller combinations.