TY - GEN
T1 - Bootstrap based pattern selection for support vector regression
AU - Kim, Dongil
AU - Cho, Sungzoon
PY - 2008
Y1 - 2008
N2 - Support Vector Machine (SVM) results in a good generalization performance by employing the Structural Risk Minimization (SRM) principle. However, one drawback is O(n 3) training time complexity. In this paper, we propose a pattern selection method designed specifically for Support Vector Regression (SVR). In SVR training, only a few patterns called support vectors are used to construct the regression model while other patterns are not used at all. The proposed method tries to select patterns which are likely to become support vectors. With multiple bootstrap samples, we estimate the likelihood of each pattern to become a support vector. The proposed method automatically determines the appropriate number of patterns selected by estimating the expected number of support vectors. Through the experiments involving twenty datasets, the proposed method resulted in the best accuracy among the competing methods.
AB - Support Vector Machine (SVM) results in a good generalization performance by employing the Structural Risk Minimization (SRM) principle. However, one drawback is O(n 3) training time complexity. In this paper, we propose a pattern selection method designed specifically for Support Vector Regression (SVR). In SVR training, only a few patterns called support vectors are used to construct the regression model while other patterns are not used at all. The proposed method tries to select patterns which are likely to become support vectors. With multiple bootstrap samples, we estimate the likelihood of each pattern to become a support vector. The proposed method automatically determines the appropriate number of patterns selected by estimating the expected number of support vectors. Through the experiments involving twenty datasets, the proposed method resulted in the best accuracy among the competing methods.
UR - http://www.scopus.com/inward/record.url?scp=44649090060&partnerID=8YFLogxK
U2 - 10.1007/978-3-540-68125-0_56
DO - 10.1007/978-3-540-68125-0_56
M3 - Conference contribution
AN - SCOPUS:44649090060
SN - 3540681248
SN - 9783540681243
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 608
EP - 615
BT - Advances in Knowledge Discovery and Data Mining - 12th Pacific-Asia Conference, PAKDD 2008, Proceedings
T2 - 12th Pacific-Asia Conference on Knowledge Discovery and Data Mining, PAKDD 2008
Y2 - 20 May 2008 through 23 May 2008
ER -