ConceptAI & ML
Depth separation
Theoretical results showing that a deep network can represent a particular function efficiently while a shallow network would require exponentially many more neurons.
Know first
Named as a prerequisite by the lectures that define this. The number is how many say so. A grey one has no page yet.
Function composition1Neural network depth1Neural networks1ReLU networks1Universal approximation1Universal approximation theorem1
Lecturers
1Connections
4The lectures that define it or assume it, and the concepts next to it.