UFC Prediction Model

Model created from my TidyTuesday Series that predicts the outcomes of UFC fights using machine learning from TidyModels.
Written in R and Developed by Andrew Couch.

The model consists of sub-models that predict individual fight components: Knockdowns, Significant Strikes, Strikes, Takedowns, and Submission attempts. The models predict each fighter's respective components and then a final model calculates the component differentials and predicts an outcome probability.

The goal of this project was to show the entire process of starting and finishing a machine learning project. This entails data scraping, data cleaning, feature engineering, modeling, and deployment/app development.