Algorithmic Aspects of Machine Learning

Calendar

SES #	TOPICS	KEY DATES
1	Introduction
2–5	Nonnegative Matrix Factorization Discussion: When does well-posedness lead to better algorithms?
6–12	Tensor Decompositions Discussion: When do algorithms rely (too much) on a distributional model?	Problem Set 1 due on session 9
13–17	Sparse Coding Discussion: When does belief propagation (provably) work?
18–22	Learning Mixture Models Discussion: Is nature an adversary? And if not, how can we model and exploit that?	Problem Set 2 due on session 20
23–26	Linear Inverse Problems Discussion: Do we have enough average-case assumptions?	Final Project due on session 26