College Football Game Prediction

Photo credit: Huffington Post

It's an age-old debate: "Your team beat us this year but if it was other year, you know we'd win, right?" Let's see.

I've collected and cleaned data from the API at collegefootballdata.com to train a classification model predicting the winner of a college football game given average season statistics since 2015. The model is then deployed so users can select any 2 teams from the time period and prove that their team in year X would actually beat their friend's team in year Y.

Overview

This project is divided into 4 parts:

1. Data Collection and Cleaning
2. Data Visualization and Exploration
3. Model Building, Evaluation, and Optimization
4. Deployment (TODO)