论文部分内容阅读
为解决辐射源识别的问题,研究了变精度粗糙集模型生成规则。由概念的正域中的对象类条件属性的描述,得到该概念的充分条件,依据知识间的依赖性作为优化规则所要考虑的因素来获取规则,由此得到的规则更加简化,且过程简单,并以雷达辐射源用途识别为例给出识别实例。实验结果表明:该方法忽略了少数实例的特殊性,对辐射源用途识别的结果与实际情况相符。该算法降低了决策规则的复杂度,能够有效地进行辐射源识别。
In order to solve the problem of radiation source identification, the rules of generating variable precision rough set models are studied. The sufficient condition of the concept is obtained from the description of the condition attributes of the object class in the positive domain of the concept and the rules are obtained according to the dependence between the knowledge as the factors to be considered in the optimization rules. The rules obtained are simplified and the process is simple. An example of recognition is given by taking radar source identification as an example. The experimental results show that this method ignores the particularity of a few examples and the result of identifying the use of radiation source is consistent with the actual situation. The algorithm reduces the complexity of decision rules and can effectively identify radiation sources.