Do Conv-nets Dream of Psychedelic Sheep?

Author: Matthew Dearing

In deep learning, understanding your model well enough to interpret its behavior will help improve model performance and reduce the black-box mystique of neural networks.

Go to Source