Given the following separable training set of labeled points in R: S = {((1,0), +1), ((2,0),...

 

Transcribed Image Text:

Given the following separable training set of labeled points in R: S = {((1,0), +1), ((2,0), +1), ((3,0),+1), ((0,1),1)} a) Perform as many iterations w(t) as needed of the Perceptron algorithm to get a separating line (assume that such a line passing through the origin exists, that is, assume there is no bias) b) What is the margin of the obtained line with respect to the given sample points? c) Based on the number of iterations you needed for convergence and the theorem on the maximum number of iterations, find an upper bound for the maximum achievable margin. Given the following separable training set of labeled points in R: S = {((1,0), +1), ((2,0), +1), ((3,0),+1), ((0,1),1)} a) Perform as many iterations w(t) as needed of the Perceptron algorithm to get a separating line (assume that such a line passing through the origin exists, that is, assume there is no bias) b) What is the margin of the obtained line with respect to the given sample points? c) Based on the number of iterations you needed for convergence and the theorem on the maximum number of iterations, find an upper bound for the maximum achievable margin.