UFC predictor is a machine learning program made in python that predicts the winner of a UFC fight. It was made as a project for Data Science class at Insper by Victor Vergara Arcoverde de A. Cavalcanti and Edgard Ortiz Neto with the assistance of Professor Fábio José Ayres and Assistant Professor Bárbara Tieko Agena.

For training we are using all the data of all fights in the UFC history. At the moment we are only using the data for Heavyweight for testing purposes. At the moment we have two models up and running, a random forest one and a logistic regression one, we still have a lot of work to do on them for them to work better, at the moment we have around 67% of accuracy on both of them.

In the final project the user will be able to write the names of 2 fighters in the same category and the model will predict the winner.

Files in directory:

Projeto3.ipynb - Jupyter file that the code is writting on.

data.xlxs - Excel file that has all the data from all UFC fights.

Link for database: https://www.kaggle.com/rajeevw/ufcdata#data.csv

If you have any doubts or suggestion be sure to send us a message via GitHub or e-mail ([email protected]) Libraries needed to run:

-sklearn;

-pandas;

-Numpy;

-matplotlib;

-statsmodels.