Multi-input fuzzy reasoning employing memory network method and global/local reasoning method
Haruhiko Arikawa, M. Mizumoto · 2002
In conventional fuzzy reasoning, when the number of input increases, the processing speed (i.e. FLIPS) decreases and the number of fuzzy rules in combination with input labels is thought to increase exponentially. This paper aims to improve these two problems by using the memory network method and global/local reasoning method.>