Self-Taught Learning to Deep Networks
From Ufldl
(Created page with "== From Self-Taught Learning to Deep Networks == Recall that in self-taught learning, we first train a sparse autoencoder on our unlabeled data. Then, given a new example <math...") |
(→From Self-Taught Learning to Deep Networks) |
||
| Line 5: | Line 5: | ||
hidden layer to extract features <math>\textstyle a</math>. This is shown as follows: | hidden layer to extract features <math>\textstyle a</math>. This is shown as follows: | ||
| - | [[ | + | [[File:STL_SparseAE_Features.png]] |
Now, we are interested in solving a classification task, where our goal is to | Now, we are interested in solving a classification task, where our goal is to | ||
| Line 18: | Line 18: | ||
follows: | follows: | ||
| - | [[ | + | [[File:STL_Logistic.png]] |
If we now look at the final classifier that we've learned, in terms | If we now look at the final classifier that we've learned, in terms | ||
| Line 25: | Line 25: | ||
particular, the final classifier looks like this: | particular, the final classifier looks like this: | ||
| - | [[ | + | [[File:STL_ComnbinedAE.png]] |
This model was trained in two stages. The first layer of weights <math>\textstyle W^{(1)}</math> | This model was trained in two stages. The first layer of weights <math>\textstyle W^{(1)}</math> | ||
