Abstract
This paper describes a new neural network modeling method for rule extraction, which incorporates an integrated learning algorithm for structure and weight parameters. Although neural networks have wide application, in a multi-layer neural network the establishment of structure parameters such as the number of units and connections between units depends on the experience of the engineer and hence the accuracy of mapping is reduced in many cases. In the proposed method, the structure and weight parameters of neural network in each problem are automatically determined by optimizing production rules by genetic operations. The optimizaiton can be quickly realized by parallel computation.
