About The Project

Look! Jon Jones only has a 32% chance of beating Stipe Miocic for the upcoming UFC 309!

A Jupyter Notebook that cleans a UFC dataset and utilizes 6 different machine learning models to predict future UFC fight outcomes. The models with their respected accuracies: Logistic Regression = 78.99%, Random Forest = 81.98%, Decision Trees = 77.06%, Naive Bayes Model = 74.94%, KNN = 79.26%, and CSM = 79.1%. The highest one being Random Forest, Hence is used as the model for the predicting function on the bottom of the notebook.

Installation

1. Clone the repo

   git clone https://github.com/Ch3mson/ufc-predictor

2. Change directory to project and ensure Python3 and pip are installed

   pip install -r requirements.txt

3. Ensure Jupyter Notebook extension is insalled on your IDE (VSCode)
4. Run notebook and adjust your desired values in the last cell. Example:

   match_probability('Conor McGregor', 'Charles Oliveira')

(back to top)

Acknowledgments

• Inspiration

(back to top)