My finished thesis, graduation, and the future

It’s been a long and wonderful four years at the University of Texas at Austin. A huge thanks to my friends, family, and teachers, all of whom did much to support me thus far. Here’s a quick overview of where I am, what I’ve done, and where I’m headed.

My thesis

I’m quite happy to announce that I’ve finished and successfully defended my undergraduate thesis, which you can now read here on this very website. The title is a bit of a mouthful, I’m afraid: An Analysis of Methods of Integration of Hyperparameter Optimization on Neuroevolutionary Models. The basic idea revolves around figuring out how, and whether, to optimize hyperparameters of the neural network training process (e.g. batch size and learning rate) when you’re automatically evolving the architectures of those neural networks over time. I make use of the concept of dataset meta-features, as popularized in the meta-learning literature. If you are really bored and for some reason want to see me explain the topic myself, well, I’ve uploaded a recording of my final presentation to YouTube. If you are genuinely that desperate for entertainment, however, I strongly urge you to read a book instead. I’ve been binging Gödel, Escher, Bach lately and it’s been fantastic so far.

Closing out my undergrad

I’m also quite happy to say that I’ve succeeded in my long-held goal of getting a dual major. Besides my B.S. in Computer Science, which I did through the Turing Scholars Honors program, I also managed to snag a B.S. in Mathematics, and I graduated today with High Honors. While my high-level plan has therefore been unchanged since high school, I think it would still be fair to say that I’ve had a lot of changes of opinion over the past few years. I originally had a strong interest in applied mathematics, financial math, and machine learning, and while I don’t particularly hate any of those fields now, it’s also true that these days I’m a lot more interested in pure math, type theory, and the like.

Perhaps you could say that unfamiliar knowledge tends to be alluring, especially compared to what is boring and well-understood. But then again, I’ve studied a fair bit of topology lately, and it doesn’t seem any less interesting to me, so maybe that’s not true at all. Regardless, it’s clear to me that there’s still plenty more for me to learn, which brings me to…

The future

I’ve decided to keep my education going just a little bit longer. This fall, I will be starting a Master’s in Computer Science at Carnegie Mellon University. I had some other options at CMU (I got into their Machine Learning program, as well as their Language Technologies Institute), but ultimately I decided that their general MSCS program was ideal for the breadth of areas and flexibility that I was looking for. Hopefully I can also find time to cover some mathematical topics that I didn’t have time to study during my undergrad, such as algebraic topology and differential geometry. I’m a bit worried that the COVID-19 pandemic will throw something of a spanner in the works of my plans, but I’m confident in my ability to make the most of the opportunities I have available.

On to the next adventure, then.