Why do we train the threshold as a weight?
I understand the process of how to do it (the bias and augmented weight vector) but do not get the importance or practical applications of doing so.
$\begingroup$
$\endgroup$
1
-
1$\begingroup$ Short answer: "Some things cannot be learned without a bias.", for longer answers see: stackoverflow.com/questions/2480650/… $\endgroup$– Lenar HoytJul 28 '15 at 11:47
Add a comment
|