Classifier Margin f est f( w, b=sign ( w x-b denotes +1 denotes -1 Define the margin of a linear classifier as the Width that the boundary could be increased by before hitting datapoint Copyright o 2001, 2003, Andrew W.Modre Support Vector Machines: Slide 9Copyright © 2001, 2003, Andrew W. Moore Support Vector Machines: Slide 9 Classifier Margin f x a y est denotes +1 denotes -1 f(x,w,b) = sign(w. x - b) Define the margin of a linear classifier as the width that the boundary could be increased by before hitting a datapoint