ali_kh_377

ممنوم بابت جوابتون.
فقط یه سوال دیگه داشتم اینه که تو قسمت پایین منظور از f(x,y) منهای f(x,guess) چیه؟
1-Let N = number of iterations through training set,
and T = the size of the training set.

2-Initialize the weight vector w to all zeros,
and the average weight vector wa to all zeros,
and the counter c to 0.

3-Iterate N times through the training set:
For each x, y in the training set:
a) Let guess = classify(x) using our current w
b) If y is not equal to guess:
•( w := w + f(x, y) - f(x, guess
•( average_weight := (N*T - c) / (N*T
•(wa := wa + average_weight * (f(x, y) - f(x, guess
c) Increase the counter c

4-Return wa