论文部分内容阅读
为了解决高炮系统防空效能仿真中战技指标的评价问题，本文提出了将神经网络学习知识转换为符号化特征表示示例，并通过分类决策树方法归纳为规则的知识自动获取方法，设计并实现了一个基于神经网络和符号归纳学习的高炮防空效能评价知识获取系统──EKANS，给出了分类决策树的链式数据结构和实现算法。系统兼有计算模型的精确性和符号模型的易理解性两大优势。目前该系统已研制成功。
In order to solve the evaluation of the tactical and technical indicators in the air defense effectiveness simulation of the antiaircraft artillery system, this paper presents an example of transforming the neural network learning knowledge into the symbolic representation of the features and the method of automatic decision making by the method of classification and decision tree. A knowledge acquisition system of air defense effectiveness evaluation based on neural network and symbolic induction learning - EKANS, a chain data structure of classification decision tree and its implementation algorithm are given. The system has both the advantages of computational model accuracy and symbolic model comprehensibility. The system has been successfully developed.