See for example combining PCA with SVM; PCA extract features which serve as inputs to an svm classifier rather than working on original data; this leads to building better models since it learns from non noisy data.
On Thursday, May 9, 2013 9:04:46 AM UTC+1, Vidin Sujith wrote: