This page is read only. You can view the source, but not change it. Ask your administrator if you think this is wrong.

====== End-to-end deep learning ====== Feasability depends on the amount of available data. Pro: * Let data speak (no forced representation) * Less hand-designing of components Cons: * Large amount of data needed * Excludes useful hand-designed features * Multi-step approach is used, when more data for subtasks is available