Courses

EECS 491: Distributed Systems ¶

In progress · Brian Noble · Rating 4/5 so far · Workload 2/5

Taking it now. Noble is great: clear lectures, good real-world stories, funny, and he cares about his students.

Projects are in Go, which was new to me, and I like it. No mutexes allowed, only goroutines and unbuffered channels. Hated that at first, then it made sense. Most of my bugs have been deadlocks.

Not a 5 because I had to rewatch the goroutine lectures a few times, and lexical confinement took me a while. Not much code to write, but a lot of thinking.

Verdict: take it. Light load, so it pairs well with a heavier class.

EECS 445: Machine Learning ¶

In progress · Jenna Wiens & Sindhu Kutty · Rating 5/5 so far · Workload 3.5/5

Taking it now. Love the topic, and the class goes deep. Wiens takes notes live and draws everything out, which works well for me.

The math is heavy. I had to catch up on linear algebra and gradients. Workload is mostly projects and homework.

It connects to my cardiac ML research both ways: the research made parts of it familiar, and the class explains why the stuff I used works.

Verdict: take it if you’re curious about ML.

EECS 376: Foundations of Computer Science ¶

Emily Graetz · Rating 2/5 · Workload 4/5

Not for me. Abstract, proof-heavy, no code. I just got through it.

The 2/5 is the workload, not the teaching. Lots of homework for what I got out of it. Some ideas are cool though, like undecidability and P vs NP.

Advice: find a study group.

EECS 485: Web Systems ¶

Andrew DeOrio · Rating 5/5 · Workload 3.5/5

My favorite class. Every project is real: Insta485, a MapReduce framework, and a search engine on top of it. More breadth than depth, which is why I liked it.

New to me: how DNS works, round-robin load balancing, AWS, and SaaS vs PaaS. The hardest part was debugging the distributed projects.

DeOrio’s lectures are clear, the course is well built, and he teaches things the way industry does them.

Verdict: take it, period. It’s good for everyone because it opens your eyes to the whole industry.

EECS 370: Computer Organization ¶

Jon Beaumont · Rating 4/5 · Workload 3/5

Very good class. Pipelining and caches clicked for me, and I think about the cache when I write code now. Clear lectures, well organized. The projects were tedious.

Advice: practice paging and virtual memory before the exam.

EECS 281: Data Structures & Algorithms ¶

Half term · Marcus Darden & David Paoletti · Rating 5/5 · Workload 4/5

Took it in the spring half term. Brutal but worth it. I preferred Darden’s lectures.

My favorite project was P1, path finding. None of them were that hard if you follow the spec. It helped me the most with interviews.

Advice: start projects on time.

EECS 280: Programming & Data Structures ¶

James Juett · Rating 4/5 · Workload 2/5

I knew some C++ already, so it was light for me, close to 1/5 if you have a background. Still worth it. The main thing I got out of it was writing real tests. Juett’s lectures are clear and the course is well organized.

Liked the Piazza classifier. Hated Euchre because I didn’t know the rules.

Advice: learn Euchre before that project.