Re: Quick clarification on the growth function
Not sure if 'location' of the point can be changed to determine the growth function. For example, in case of N=4 with 2D perceptron, growth function is 14 because the perceptron cannot separate diagonally opposite points of the same class. If we change the position of the points in this case we would get growth function as 16. It seems like one can get at least one arrangement of all N points that can be shattered for any N by placing the points at appropriate position. I guess I am missing something.
|