Support vector machine for HRRP classification

Radar target identification schemes by using high resolution range profile(HRRP) as features have been studied extensively. In practical systems we usually have only a very limited amount of training data. Therefore how to train a classifier with good generalization performance based on the training set is obviously a challenging task. This paper introduce the newest branch of statistic learning theory, support vector machine(SVM) to range profile classification. The range profiles of two targets were classified by SVM and LVQ (Learning Vector Quantization). Experiment results show that applying SVM to range profiles classification can get higher correct classification rate and better generalization performance.