We propose a simple use of principal component analysis in feature space that allows the derivation of optimal predictive kernel regression. The proposed approach is shown to perform well on both artificial and real data. Despite its incredible simplicity, the proposed method is found to compete very well with sophisticated statistical approaches like the Relevance Vector Machine and the Support Vector Machine.
Department, Program, or Center
The John D. Hromi Center for Quality and Applied Statistics (KGCOE)
Fokoue, Ernest, "Optimal predictive kernel regression via feature space principle components" (2011). p. 87-108. Accessed from
RIT – Main Campus