How to determine the coefficient of svm classifiers for linear kernels in R? -
i using kernlab package svm in r.i using linear kernel can directly check importance of feature vectors, variables.using coefficients of these feature vectors,i required calculate weight of various factors in model,so linear separating plane svm draw in feature space can evaluated. want calculate w in transpose(w)*x + b. please suggest done. used fields alpha , b , apha index , tried logically calculate weight vector, verify if calculating correctly tried predict on test sample correct predicted score, , did not match value predicted inbuilt predict function. how calculate weights?
Comments
Post a Comment