一个基于神经网络的专家系统

医疗专家系统主要使用基于知识的技术，其中的决策规则和策略来自于人类的专家。把这些知识和各种推理方法结合，可以建立一个模拟专家决策过程的系统。建立这样一个系统，需要经常与专家磋商，以获取专家的知识，因而需要大量的时间和精力。为此，本文提出直接从数据中提取有效的信息，即用神经网络提取隐含在大量数据中对医疗诊断有效的信息，继之与基于规则的知识，各种推理方法相结合，建立一个神经网络专家系统。

An Expert System Based On Neural Networks

Expert systems of medicine have relied heavily upon knowledge-based techniques, in which decision making rules and strategies are derived through consultation with experts, These techniques,coupled with many kinds of reasoning methods,may produce the systems which model the human decision making process. This approach requires extensive interviewing of experts to extract the knowledge from experts, so it consumes much time. The approach of extracting available information directly from data bases, i. e. A neural network model used to extract the information available to medical diagnosis from data bases, is presened. Then the neural network is conjuncted with rule-based knowledge, incorporating many kinds of reasoning methods.