Optimization
Learning is finding the lowest point of an error surface. This is how you get there.
Understand first: Calculus · Linear Algebra
The planned roster
This is the intended sequence, first principles first. It is the spec a contributor builds against, and it is the reason this domain is greyed out instead of guessed at.
-
1
what optimization is
-
2
critical points
-
3
convexity
-
4
gradient descent
-
5
the learning rate
-
6
stochastic gradient descent
-
7
local vs global minima
Missing a topic?
This roster is a proposal, not a decree. If something belongs in Optimization and is not listed, ask for it. Requests are public, and anyone can pick one up.