用神经网络来实现基于范例的推理系统

Building Case-based Reasoning System with Neural Networks

范例推理与神经网络有一种自然的联系。神经网络有许多优点,利用神经网络来实现范例推理可以取得非常好的效果。文章首先详细探讨了在范例推理中使用的神经网络模型与技术,并给出了其上的搜索与学习算法以及数据挖掘算法,旨在提高范例推理系统的鲁棒性和知识获取的自动化程度。

There is a natural relationship between case-based reasoning and neural network. This paper examines the possibility of using neural network as a method of building such case-based systems. Better results in CBR system could be achieved by using neural network. A simple efficient case-based system structure is constructed with neural network. Also, some algorithms are proposed and tested. The purpose of research is to heighten the robustness and automation of knowledge acquisition in case-based system. ;;;;;