Hello,
I have some trouble understanding the solution of question 3.b concerning the application of the perceptron algorithm, from last weeks exercises. Why is it that when calculating the output for each point, -1 is added as a third coordinate to every point ? For instance, the output of the first point (2,-1) will be calculated by applying the heaviside function to (1,0,0).(2,-1,-1). I understand that we must add a third coordinate to the points when doing the scalar product with the weights because of the threshold of the perceptron adding a third parameter. But why should it be -1 ? Does it matter ? How do we find which value to use ?
