October 2017
Beginner to intermediate
270 pages
7h
English
As we have observed in the previous chapters, overfitting is a potential problem for every model. This is also the case for neural networks, where data can do very well on the training set but not on the test set, which renders it useless for generalization.
For this reason, in 2012, a team led by Geoffrey Hinton published a paper in which the dropout operation was described. Its operation is simple:
Read now
Unlock full access