willieodonnel
-
Quantifying Positional Difficulty With a Heuristical Approach
This was a paper which I wrote as an extension of my research on classifying tactically interesting positions. Unfortunately I did not finish the project as I got quite sick at the end and after working for roughly 4 months between the two projects for probably 8 hours a day, I was done (also, towards…
-
Finetuning an LLM for Disproving Rumors Related to FEMA
This was a project I did for my PSTAT 134 class my last quarter at UCSB. Together with my groupmates Denis Lomov, Sanchit Mehrotra, and Netasha Pizano, we finetuned a version of GPT 2 on Q&A pairs found on FEMAs website so that it could answer questions related to FEMA and common misconceptions about what…
-
Neural Network-based Pipeline for Identifying Tactically-Interesting Chess Positions
This was the beginning of my 2 part project identifying tactically-interest positions using ML. It was extremely fun and made me realize how much I enjoy designing pipelines of this nature. I was working on this project alongside my friend Ayush Nair who provided us with invaluable expertise and access to his GPU. One thing…
-
Building a Model for Predicting Professional CS:GO Match Outcomes (R)
This was my first ever machine learning project for my introduction to statistical machine learning class at UCSB. CS:GO was a favorite game of mine for years and while I hate the idea of gambling, it felt like an exciting opportunity to explore an area where I could develop something that might be able to…
William O'Donnell 