Contents

Curriculum

part 1Sessions 1–7

Foundations

Setup, what AI is, what a contest problem looks like, and Python for data.

S1Setup and what AI issession 1lesson · ~2h
S2Intro to contest AIsession 2lesson · ~2h
S3Python for datasession 3–5lesson · ~3 weeks
Full pipeline: once you have the Python basics, one run from reading the data to a submission, end to end.
S6EDA and preprocessingsession 6–7lesson · ~2 weeks

part 2Sessions 8–17

Classic machine learning

The classic models and honest evaluation, plus search, NLP and vision without neural nets.

part 3Sessions 18–24

Deep learning

Neural networks in PyTorch, CNNs, embeddings and reinforcement learning.

note

Over the winter break there's one optional task: two archive problems, no hard deadline.