tailieunhanh - Báo cáo hóa học: " An Improved Way to Make Large-Scale SVR Learning Practical"

Tuyển tập báo cáo các nghiên cứu khoa học quốc tế ngành hóa học dành cho các bạn yêu hóa học tham khảo đề tài: An Improved Way to Make Large-Scale SVR Learning Practical | EURASIP Journal on Applied Signal Processing 2004 8 1135-1141 2004 Hindawi Publishing Corporation An Improved Way to Make Large-Scale SVR Learning Practical Quan Yong Institute of Image Processing and Pattern Recognition Shanghai Jiao Tong University Shanghai 200030 China Email quanysjtu@ Yang Jie Institute of Image Processing and Pattern Recognition Shanghai Jiao Tong University Shanghai 200030 China Email jieyang@ Yao Lixiu Institute of Image Processing and Pattern Recognition Shanghai Jiao Tong University Shanghai 200030 China Email lxyao@ Ye Chenzhou Institute of Image Processing and Pattern Recognition Shanghai Jiao Tong University Shanghai 200030 China Email Received 31 May 2003 Revised 9 November 2003 Recommended for Publication by John Sorensen We first put forward a new algorithm of reduced support vector regression RSVR and adopt a new approach to make a similar mathematical form as that of support vector classification. Then we describe a fast training algorithm for simplified support vector regression sequential minimal optimization SMO which was used to train SVM before. Experiments prove that this new method converges considerably faster than other methods that require the presence of a substantial amount of the data in memory. Keywords and phrases RSVR SVM sequential minimal optimization. 1. INTRODUCTION In the last few years there has been a surge of interest in support vector machine SVM 1 . SVM has empirically been shown to give good generalization performance on a wide variety of problems. However the use of SVM is still limited to a small group of researchers. One possible reason is that training algorithms for SVM are slow especially for large problems. Another explanation is that SVM training algorithms are complex subtle and sometimes difficult to implement. In 1997 a theorem 2 was proved that introduced a whole new family of SVM training procedures. In a nutshell Osuna s theorem showed