# Giant pile of ML problems

**Links for download: ****pdf**** (****code****)**

*Last updated: January 2023*

I wrote most of these questions from scratch (except standard/famous questions), so while I'm happy to publicize and share, I do appreciate a citation

@misc{giant_pile_problems,

author = "Yifan Sun",

title = "Giant Pile of ML Problems",

howpublished = "Problem set",

month = "January",

year = "2023",

}

**Breakdown**:

Conceptual questions: used to determine if people are paying attention / understanding the purpose behind the math

Conceptual short answer questions: generates some discussion and thinking about the "bigger picture ML issues"

Basic probability (undergrad level)

Advanced probability (grad level)

Graphical models

Basic linear algebra (undergrad level)

Advanced linear algebra (grad level)

ML models: mostly for coding up ML models from scratch (e.g. the science behind sci-kit learn) and exploring their strengths and weaknesses

Ensemble methods: includes toy examples for building intuition, and full coding problems.

Generalization: Simple toy problems to exercise thinking about generalization issues.

Optimization analysis (background, grad level)

Optimization methods (just a few for the devoted student)

Machine learning theory (still working on beefing up this section. Intentionally simplistic, good for undergrad or grad depending on the problem.)

Deep learning (just a few simple practice questions to establish fundamentals, definitely not the emphasis.)

I appreciate feedback / notes on corrections!