Foundations 0/0

Foundations

Neural networks

A network as layers of weighted sums and nonlinear activations, trained by backpropagation with an optimizer such as SGD with momentum or Adam. Every model in the later tracks is trained this way.

0 / 10 solved

Questions

0/10 solved
Question Module Tags Difficulty Success
1. Forward March, Backward Blame NN Fundamentals & Backprop
forward-passneural-networks
Easy 85%
2. Chain Reaction NN Fundamentals & Backprop
backpropagationchain-rule +4
Medium 90%
3. Stacking the Deck NN Fundamentals & Backprop
depthrepresentation-learning
Medium 100%
4. Goldilocks Weights NN Fundamentals & Backprop
weight-initializationvanishing-gradients
Medium 80%
5. Dead End Street Activations & Optimization
activationsrelu
Medium 100%
6. The Slow Roll Activations & Optimization
optimizationmomentum
Medium 100%
7. First Step Is Alpha Activations & Optimization
adambias-correction +2
Medium 69%
8. Warm Start, Cool Finish Activations & Optimization
learning-rategradient-clipping
Medium —
9. Where The Decay Goes Activations & Optimization
adamwweight-decay +2
Hard —
10. What objective does SAM approximately optimize? Activations & Optimization
Machine LearningTraining & Optimization
Medium —