A Fast Training Algorithm for SVM and Its Application in HRRP Classification

By choosing the most informative patterns that have the most possibility to become the support vectors in the training data by using the convex hulls algorithm, a fast training algorithm for SVM (QhullSVM) is given in this paper. Experimental results reveal that the given QhullSVM has better training performance comparing with the traditional training algorithm for SVM, and has distinct performance improvement when deal with the low dimension and large size dataset. By applying the QhullSVM to the training of SVM in the IDDAGSVM, an improved algorithm DDAGQSVM obtained. After analyzing the characteristics of the high range resolution profile, and using the principal component analysis to decrease the dimensionality of the sample set, a scheme for HRRP classification by DDAGQSVM was given. The effectiveness of the given scheme for HRRP classification by DDAGQSVM was testified by the measured data for 5 targets. Experimental results reveal the effectiveness of the algorithms in this paper.