ConceptAI & ML
Residual network
The final concept needed before introducing algorithms, dealing with the capacity of a cut in flow networks.
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.
Activation function1Augmenting paths algorithm1Convolutional neural networks1Cuts in flow networks1Gradient descent1Graph theory1Max flow problem1Max-flow min-cut theorem1
Opens up
Concepts that name this one as a prerequisite.
Lecturers
4Each in their own terms, from the easiest level.
Fei-Fei Li, Justin Johnson & Serena Yeung
A deep convolutional architecture featuring shortcut connections that form residual blocks, allowing layers to learn identity functions and easing gradient flow.Peter Shor
A network that shows the remaining capacities on edges and allows pushing flow backwards along edges with existing flow.Chris Ré
A neural network architecture that uses residual connections to model the difference between the input and the target, facilitating deeper networks.
Lectures that assume it
2Connections
14The lectures that define it or assume it, and the concepts next to it.
Concepts8
- Activation function
- Augmenting paths algorithm
- Convolutional neural networks
- Gradient descent
- FractalNet
- Augmenting path