Support vector machine, and its generalization, is a * very * hot area for machine learning people.
You may want to check out the site http://www.kernel-machines.org/ for more information on SVM. In particular, there is a tutorial on SVM. Note that "kernel" in SVM literature means something * completely * different from "kernel" in non-parametric regression. One research direction for SVM is to derive traditional statistical procedures using the kernel trick. Examples include principal component analysis, Fisher's discriminant analysis, k-means clustering, and canonical correlation. This may be the reason why you see SVM and canonical correlation coming together. deepak <[EMAIL PROTECTED]> wrote: > Please tell me the relation between SVM and Canonical correlation and > other statistical concepts (with examples if possible) > I don't have any idea of Support Vector Machines but I came across in > the multivariate statistics. > thanks in advance > deepak . . ================================================================= Instructions for joining and leaving this list, remarks about the problem of INAPPROPRIATE MESSAGES, and archives are available at: . http://jse.stat.ncsu.edu/ . =================================================================
