SOARES, Carlos, Pavel BRAZDIL and Petr KUBA. A meta-learning method to select the kernel width in Support Vector Regression. Machine Learning Journal. Netherlands: Kluwer Academic Publishers, 2004, vol. 54, No 3, p. 195-209. ISSN 0885-6125. |
Other formats:
BibTeX
LaTeX
RIS
@article{555356, author = {Soares, Carlos and Brazdil, Pavel and Kuba, Petr}, article_location = {Netherlands}, article_number = {3}, keywords = {meta-learning; parameter setting; support vector machines; Gaussian kernel; learning rankings}, language = {eng}, issn = {0885-6125}, journal = {Machine Learning Journal}, title = {A meta-learning method to select the kernel width in Support Vector Regression}, url = {http://ipsapp009.kluweronline.com/IPS/frames/toc.aspx?J=4984&I=140}, volume = {54}, year = {2004} }
TY - JOUR ID - 555356 AU - Soares, Carlos - Brazdil, Pavel - Kuba, Petr PY - 2004 TI - A meta-learning method to select the kernel width in Support Vector Regression JF - Machine Learning Journal VL - 54 IS - 3 SP - 195-209 EP - 195-209 PB - Kluwer Academic Publishers SN - 08856125 KW - meta-learning KW - parameter setting KW - support vector machines KW - Gaussian kernel KW - learning rankings UR - http://ipsapp009.kluweronline.com/IPS/frames/toc.aspx?J=4984&I=140 N2 - The Support Vector Machine algorithm is sensitive to the choice of parameter settings. If these are not set correctly, the algorithm may have a substandard performance. Suggesting a good setting is thus an important problem. We propose a meta-learning methodology for this purpose and exploit information about the past performance of different settings. The methodology is applied to set the width of the Gaussian kernel. We carry out an extensive empirical evaluation, including comparisons with other methods (fixed default ranking; selection based on cross-validation and a heuristic method commonly used to set the width of the SVM kernel). We show that our methodology can select settings with low error while providing significant savings in time. Further work should be carried out to see how the methodology could be adapted to different parameter setting tasks. ER -
SOARES, Carlos, Pavel BRAZDIL and Petr KUBA. A meta-learning method to select the kernel width in Support Vector Regression. \textit{Machine Learning Journal}. Netherlands: Kluwer Academic Publishers, 2004, vol.~54, No~3, p.~195-209. ISSN~0885-6125.
|