6.034 Artificial Intelligence - Recitations, fall 2004 online slides on learning

Next: Next Layer Previous: Delta Rule

Hidden-to-output Weights: updating

$\Delta_i$ = \(Err_i * g'(in_i)\) or

$\Delta_i$ = \((t_i - o_i) * g'(in_i)\), where

$t_i$ = true / target output for node i

$o_i$ = actual / calculated output for node i (t - o is error for node i)

$in_i$ = sum of inputs

$g'_i$ = derivative of transfer function