- A Project to predict the outcome of a UFC fight between 2 fighters.
- Combined 2 datasets. Fight details which had numerical as well as categorical features of the fight statistics between 2 fighters. And fighter details of each figher which had the historical statistics of the fighter leading up to the fight.
-> Did data analysis on both the datasets separately and also after combining them.
-> Cleaned the data to remove noise from features.
-> Performed feature engineering to extract more important features like fight duration, round duration etc.
-> Encoded categorical features and normalized the numerical features
-> Did K-fold cross-validation on Random Forest Classifier and XGB classifier and achieved 86.18% and 90.08% accuracy score respectively