Generating Weighted Fuzzy Production Rules using Neural Networks
Tiegang Fan, Shu-tian Wang, Jun-Min Chen · 2006
Weighted fuzzy production rules enhance the knowledge representation power of rule. This paper proposes a way to generate weighted fuzzy production rules using neural networks. First the knowledge in the data is transformed into neural network. Through analysis of the weights of the neural network, a matrix of importance index is constructed. Then weighted fuzzy production rules are extracted from the neural network. In order to reflect the knowledge implied in the neural network accurately, a corresponding reasoning algorithm is constructed. The effective of the approach is demonstrated by the experiment