Support Vector Regression with ANOVA Decomposition Kernels
Mark O. Stitson, Alex Gammerman, Vladimir N. Vapnik, Volodya Vovk, Chris Watkins, Jason Weston · The MIT Press eBooks · 1998
Support Vector Machines using ANOVA Decomposition Kernels (SVAD) [Vapng] are a way of imposing a structure on multi-dimensional kernels which are generated as the tensor product of one-dimensional kernels. This gives more accurate control over the capacity of the learning machine (VCdimension) . SVAD uses ideas from ANOVA decomposition methods and extends them to generate kernels which directly implement these ideas. SVAD is used with spline kernels and results show that SVAD performs better than the respective non ANOVA decomposition kernel. The Boston housing data set from UCI has been tested on Bagging [Bre94] and Support Vector methods before [DBK + 97] and these results are compared to the SVAD method. 1 Introduction In this paper we will introduce ANOVA kernels for support vector machines. We firstly introduce multiplicative kernels, which form the basis of the ANOVA kernels, then we introduce the general ANOVA decomposition idea. From this we derive ANOVA kernels and lastly sh...