1

I have seen in many areas people scale their data before passing to a classifier or learning algorithm . I noticed that by doing that classifier would be really fast. For instance in python X=preprocessing.scale(X) is used before doing svm.SVC().fit(X,y) . Does this have any effect on classifier?

Edit : X=preprocessing.scale(X) is supposed to

Standardize a dataset along any axis Center to the mean and component wise scale to unit variance.

Does scaling to unit variance have any effect?

MysticForce
  • 121
  • 4
  • See here: http://stats.stackexchange.com/questions/65094/why-scaling-is-important-for-the-linear-svm-classification or here: http://stackoverflow.com/questions/15436367/svm-scaling-input-values or here: http://stats.stackexchange.com/questions/77876/why-would-scaling-features-decrease-svm-performance/77894#77894 If those don't help then please elaborate on what specifically you don't understand. – ilanman Jan 10 '17 at 11:40

0 Answers0