基于神经网络的知识获取

本文提出了用基于规则专家系统与神经网络的集成，该系统实现了从实例中自动获取知识的功能．在产生和控制不完全情况方面提高了专家系统的推理能力．它使用无导师学习算法的神经网络来获取正规数据，并用一个符号生成器把这些正规的数据变换成规则．生成规则和训练后的神经网络作为知识库嵌于专家系统中．在诊断阶段，为了诊断不明情况，可同时使用知识库和人类专家的知识，而且系统可以利用训练过的神经网络的综合能力进行诊断，并使不相符数据完整化．

Knowledge Aequisition Using Network

Is this paper, we present the integration of neural networks with a rule based expert system. The system realizes the antomatic acquistion of knowledge out of a set examples. It enhsnces the reasoning capabilities of classical expert systems with the abilities of generalize and the handling of incomplete cases. It uses neural nets with unsupervised learning algorithms to extract regularies out of case data. The generated rules and the trained nets are embeded into the expert system as knowledge bases. In the diagnosis phase, it is possible to use these konwledge bases together with human esperts' knowledge has in order to diagnoses a unknown case, furthermore the system is able to diagnoses and complete inconsistent data using the trained neural nets exploiting their ability to generalise.