摘要
通过对以往战场舰艇意图识别经验的学习,采用信息熵理论建立舰艇意图识别决策树。通过对舰艇意图识别历史数据的分析,将识别错误的元组更正后加入历史信息库,重新建立决策树,使该方法具有一定的自学能力,并通过实例进行仿真分析。结果表明,该方法具有一定的自学能力,意图识别的准确率会随着决策树的学习逐渐提高,在一定程度上克服战场目标意图的欺骗性。
Decision trees can be created by using information entropy through the research on vessel intention recognition experience. Through analyzing the intention recognition historical data, correct the error meta-groups and add them to the historical information base and reconstruct decision tree. The method has self-study ability and can carry out simulation analyze based on example. The results show that the method has self-study ability, the intention recognition accuracy rate will improve by the learning and overcome the cheating of battlefield target intention in some degree.
出处
《兵工自动化》
2010年第6期44-46,53,共4页
Ordnance Industry Automation
关键词
信息熵
决策树
海战场
意图识别
Information entropy
Decision tree
Naval battlefield
Intention recognition