Neural networks in generalizing expert knowledge
Wang, Shouhong
Neural networks in generalizing expert knowledge
In a rule-based expert system, knowledge is represented as individual rules. Rules which have the same cause and effect structure could be generalized using a mapping function in order to make the knowledge base more economic in size and more flexible in reasoning. Such a function can be generated using the monotonic back propagation neural networks. In principle, the neural networks representing the mapping function for the expert system can be substituted for all of the production rules and thus constitute a knowledge base. Compared with the traditional production rule based expert systems, neural network based expert systems are more effective in approximate reasoning
NEURAL NETWORKS
EXPERT KNOWLEDGE
H004.COM
Neural networks in generalizing expert knowledge
In a rule-based expert system, knowledge is represented as individual rules. Rules which have the same cause and effect structure could be generalized using a mapping function in order to make the knowledge base more economic in size and more flexible in reasoning. Such a function can be generated using the monotonic back propagation neural networks. In principle, the neural networks representing the mapping function for the expert system can be substituted for all of the production rules and thus constitute a knowledge base. Compared with the traditional production rule based expert systems, neural network based expert systems are more effective in approximate reasoning
NEURAL NETWORKS
EXPERT KNOWLEDGE
H004.COM