基于粗糙集支持向量机的空袭目标识别

摘要

In order to improve the ability of an air raid target recognition, a method of an air raid target recognition based on rough set support vector machine was proposed. The RS-method is advantage at processing big amount of data and removing redundancy information. So we take this method as the head system of the SVM data processing system to reduce SVM training data, therefore overcame the disadvantage of the SVM which is slow at processing big amount of data. According to the request of the recognition, this paper establishes the recognition model of "one-against-one" approach on the arithmetic of RS-SVM. The results of emulation experiments show that the method has higher recognition probability and was great superior in the an air raid target recognition.%为提高对空袭目标的识别能力,提出了一种基于粗糙集支持向量机的空袭目标识别方法.该模型用RS方法构建SVM数据处理系统的前置系统,充分利用RS理论在处理大数据量,消除冗余信息等方面的优势,减少了SVM训练数据,克服了SVM算法因为数据量太大而导致处理速度慢的缺点.根据分类识别的要求,在RS-SVM两类分类算法的基础上,建立了成对分类目标识别模型.通过仿真试验证明,该方法具有较高的识别率,在空袭目标识别中体现极强的优势.