ANFIS-SNNS: Adaptive Network Fuzzy Inference System in the Stuttgart Neural Network Simulator
Kais Brahim, Andreas Zell · 1994
In this paper the Neuro-Fuzzy system ANFIS (Adaptive Network Fuzzy Inference System) and its integration in the Stuttgart Neural Network Simulator (SNNS) is described. The rule-based knowledge base of a fuzzy system is directly mapped to the network structure of a neural network. With a hybrid learning algorithm the system adapts itself to the environment by using examples to optimize the rules. The structured network architecture also gives the possibility to extract the optimized fuzzy rules from the network after training.