Authors: Jingxiang Zhang Shitong Wang
Publish Date: 2015/06/23
Volume: 27, Issue: 6, Pages: 1717-1730
Abstract
The leaveoneout crossvalidation is an important parameter selection strategy for SVMlike family including SVM and SVR However due to the high computational complexity the adaptability of this strategy is restricted In this paper aiming at its practical application a fast leaveoneout crossvalidation method by using an adjustment factor is proposed which focusses especially on the practicability for the SVMlike family where the decision function can be expressed by explicit dotproduct of each training sample pair The ability of the proposed method in fast parameter selection is better than that of original leaveoneout crossvalidation with the same or comparable learning performance The simulation results indicate the effectiveness and speedup of the proposed leaveoneout crossvalidation methodThis work is supported by the National Natural Science Foundation of China Nos 61202311 61300151 the Natural Science Foundation of Jiangsu Province under Grant BK201221834 the Fundamental Research Funds for the Central Universities No JUDCF13031 and 2013 Postgraduate Students Creative Research Fund of Jiangsu Province No CXLX13 748 the Ministry of Education Research of Social Sciences Youth funded projects 14YJCZH206
Keywords: