# Fine-tuning Stacked AEs

 Revision as of 01:28, 22 April 2011 (view source)Watsuen (Talk | contribs) (→Recap of the Backpropagation Algorithm)← Older edit Revision as of 01:32, 22 April 2011 (view source)Watsuen (Talk | contribs) (→Recap of the Backpropagation Algorithm)Newer edit → Line 16: Line 16: = - (\nabla_{a^{n_l}}J) \bullet f'(z^{(n_l)}) = - (\nabla_{a^{n_l}}J) \bullet f'(z^{(n_l)}) \end{align}[/itex] \end{align}[/itex] + For the softmax layer, we have $\delta^{n_l} = \theta^T(I-P)$ where $I$ is the input labels and math>P[/itex] is the predicted labels. : 3. For $\textstyle l = n_l-1, n_l-2, n_l-3, \ldots, 2$ : 3. For $\textstyle l = n_l-1, n_l-2, n_l-3, \ldots, 2$ ::Set ::Set