摘要
贝叶斯分类器是使错误分类概率最小的最优方法,但必须具备先验知识,计算量也很大,从而增加了实时应用的复杂性。提出基于贝叶斯网络海上目标识别,结合贝叶斯网络对不确定事件强的推理作用,以及贝叶斯理论的数学基础,应用图形模式,使得计算量大大简化,降低了实用的复杂性。
A classify tool based on hayes theory is best way that make minimal error. If want to use this way, must master enough prior knowledge. That waster lots of time to calculate. This paper thinks out a good idea to identify an object on sea based on hayes network. It is no bad way because of inference and express method of unceitain knowledge as well as graphics mode. So this method uses less time to get result.
出处
《微机发展》
2005年第10期152-154,共3页
Microcomputer Development
关键词
贝叶斯网络
不确定推理
目标识别
hayes network
inference method of unceitain knowledge
identify object