# Fine-tuning Stacked AEs

### From Ufldl

(→Recap of the Backpropagation Algorithm) |
(→Recap of the Backpropagation Algorithm) |
||

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}</math> | \end{align}</math> | ||

- | ::For the softmax layer, we have <math>\delta^{n_l} = \theta^T(I-P)</math> where <math>I</math> is the input labels and math>P</math> is the predicted labels. | + | ::For the softmax layer, we have <math>\delta^{n_l} = \theta^T(I-P)</math> where <math>I</math> is the input labels and <math>P</math> is the predicted labels. |

: 3. For <math>\textstyle l = n_l-1, n_l-2, n_l-3, \ldots, 2</math> | : 3. For <math>\textstyle l = n_l-1, n_l-2, n_l-3, \ldots, 2</math> | ||

::Set | ::Set |