School Projects

Here are projects I completed in my courses at UGA in my Statistics/MIS classes. The projects shown here use Python and Java.

Data Science Competition

My team fit logistic regression and gradient boosted models on five factors derived using factor analysis (mimicking credit score) to minimize the amount of loans given to faulty customers. We won the competition due to our codebase, report, and presentation to Wells Fargo execs.

Senior Statistics Capstone

My team of three joined with an organization called Elephants for Africa to create a noninvasive way of tracking elephants in Zimbabwe by analyzing their footprints. We built a Convolutional Neural Network to identify which footprint belonged to each elephant. Due to the lack of data, we had to use semi-supervised methods.

Blackjack Game

I created a full-functioning blackjack game for my final project in a computer science course to demonstrate Object Oriented Programming skills in Java. The game has an interface that deals two cards and follows all the rules of normal Blackjack. It will also shuffle the deck after every turn to ensure the game is as realistic as possible.

Prophet Package Demo

For Advanced Data Management and Analytics (MIS5730 - Python), my partner and I created a demo for the time series forecasting tool Prophet.

Georgia Tech Projects

Since I am still completing the program, I do not want to make my repositories public for academic dishonsety reasons. If you would like to see my work, please contact me using the form or sent me an email at My projects include Search algorithms (A*, Dijkstra's Algorithm, etc), Bayesian Networks, Game AI (Minimax, Alpha-Beta Pruning), ML Tree Algorithms from scratch, EM Algorithms, TD-Lamda, DQN, Multi-Agent GFootball (PPO, A2C, QMIX), AI for Robotics algorithms (PID, Kalman Filters, SLAM), FF Neural Networks from Scratch, CNNs from scratch, Multi-head attention networks from scratch.

Personal Projects

Here are some projects that I completed outside of school. I actively update the college football code base. The other two projects were completed during undergrad.

Bayesball Project

This project focuses on using Bayesian Inference to accurately show the performance of several MLB metrics. For the first step, I used a Bayesian Hierarchical Linear Model to predict adjusted ERA for each team and rank the teams based on whether they are underperforming or outperforming the predicted baseline. In the future, I'd like to automate this model and use it to predict win percentages. (Beta)

College Football Handicapping

This is a project I started because I entered a college football pool and wanted to use machine learning to my advantage. I built out a full data pipeline, reading directly from the cfbd api. I build several features and fit an XGBoost model to predict what the spread of a game should be. My model is 55.7% against the spread this year, which leads the pool.

Predicting Golf Tournament Outcomes

This project incorporates statistical and machine learning models to analyze past golf data and how it can predict the upcoming tournament. Currently, the model predicts FanDuel and Draftkings points by strokes gained categories. The goal for this project is to use more predictors and similuate each tournament using MCMC to predict who will perform well.