Fall 2024 Quarter Class Review

Published:

Wow, has it really been a year since I started the program? Summary: Embracing the feeling of being stuck

Courses I took:

  • A ☆ ME 564: Mechanical Engineering Analysis
  • A ☆ Psych 509: Core Concepts in Computational Cognitive and Neural Modeling
  • Cr Psych 542: Advances in Behavioral Neuroscience
  • Cr Psych 555: Seminar in Cognition and Perception
  • Cr Psych 560: Research Strategies

Other Commitments: Various research projects and exploratory ideas.


Retrospective

Grad Student Life

This quarter was rough and demotivating at times.

Research

I faced a ton of setbacks in my research—which was bound to happen eventually—but it was still tough to deal with. I honestly believe that being a Ph.D. student isn’t (just) about being smart (honestly, I don’t think I’m that smart at all); it’s about staying optimistic and persisting against all odds. In a way, doing a Ph.D. feels borderline hubristic—do you really have what it takes to add even a tiny bit of new knowledge to everything humanity has already accumulated?

Anyway, I’m glad the roadblock and demotivation hit this quarter and not in my final year when I’d panic and probably have a breakdown. My advisor has been incredibly supportive, and that’s made all the difference.

Persistent Homology

This was pretty fun. While stuck on my main research project, I pitched an idea to my advisor: What if we use persistent homology to answer questions about our datasets, flies, or brains in space? One downside of most neuroscience research—at least research that treats the brain as a connectivity matrix—is that it misses a lot of nuance. Representing the brain as a DAG just doesn’t capture the complexity and dependeing on the question you’re asking, you’re going to be missing a lot of nuance.

It was cool to return to this because my advisor have thought about these topics before. During my lab interview way back when, we even briefly discussed topology and category theory (as much as two outsiders can) and the potential role those fields could play in neuroscience.

I might have a collaborator in this area once I’ve fleshed my research idea out more, so stay tuned for that!


Miscellaneous

Having worked in industry before grad school has been helpful. I know I have value outside my research output, which has really helped me put things into perspective.

It’s been nice that SBP opened a more accessible location. Even though my overall commute time from the gym to campus hasn’t changed much, the mental load of catching buses and waiting around has decreased significantly.


Classes

Now, on to what you’re actually here for: class reviews.

Psych 509 was a surprise for me. I went in expecting to know almost everything, but I learned quite a bit. I found it surprising how technical human psychology research can get, and I really enjoyed the class. You can tell Andrea truly cares about the topics and loves teaching. We covered reinforcement learning, drift diffusion models, deep learning, and deep RL (presented by yours truly). If you’re into theory-of-mind topics and connecting brain functions to algorithms, I highly recommend this class. It’s not particularly relevant to my research, but it gave me a lot to think about.

I’m on the fence about ME 564. The online lecture series on differential equations and dynamical systems covered the material so well that I stopped attending class or watching lecture recordings. Honestly, there was little incentive for me to schlepp myself in. It doesn’t help that the class was located far away from my office, and by the time I remembered to start walking over, class had already started. That said, I loved the lectures I did attend. Michelle and Steve clearly care about students understanding the material. The homework assignments were surprisingly easy, especially since I hadn’t done a Taylor series expansion or general engineering math in about eight years. I took this course because I feel dynamical systems will become hugely relevant in neuroscience in the coming years, and I still think that’s true.