February 2025

I competed alongside 3 teammates in Rice Datathon 2025. We had a total of 36 hours to complete a data science based project on one of 3 tracks. We chose the neuroscience track and were tasked with creating a model that would predict for different psychiatric disorders based on EEG data and minimal patient data. This task demanded both domain knowledge and statistical modeling skills. We used both Python and R to clean data and attempted some feature engineering. We attempted to use a variety of models before settling on XGBoost as our best option given the data. One of my biggest takeaways from this experience was how outside priorities influence model decisions.