A Comparative Study of Supervised Machine Learning based Models for Winner Prediction in Mixed Martial Art
The Mixed martial art is a popular sport around the world. And in mixed martial art sport, the Ultimate Fighting Championship (UFC) is one of the most famous flourishing organization. The mixed martial art as the name suggests is a combination of many fighting techniques like boxing, wrestling, jiu-jitsu, and many others. The ufcstats.com is the official website where the data related to all UFC fights from 1993 to till date can be found. The website consists of several fighter characteristics in detail. In this paper, we try to predict the accuracy of different supervised machine learning algorithms for the forecast of the winner based on a fighter’s various characteristics is done. The effectiveness of several supervised machine learning classifier based on Decision Tree, Perceptron, Random Forest(RF), Stochastic Gradient Descent (SGD), Bayes Classifier, Support Vector Machine (SVM), K-Nearest Neighbor (KNN) and eXtreme Gradient Boosting (XGBoost) classifiers are tested on a time series data of a fighter’s several features. An accuracy of up to 68% is achieved with these models.