Predicting the Winner of a Tennis Match Using Machine Learning Techniques

Winning is the primary goal of any sport. Predicting the winner of the match in advance has gained a lot of attention by sports organizations and potential bidders as it involves a lot of time and money invested in it. Now-a-days, sports organizations realize the value of data and the science in the data which can be used as an advantage to players, coaches also the potential bidders using machine learning techniques. Tennis is a challenging and unpredictable sport, yet the most exciting sport which is enjoyed by fans from all over the world. Machine learning techniques have helped in predicting the outcomes of tennis matches using various attributes. Removing the irrelevant attribute plays a very important role in getting high accuracy hence in this research we have used the Principle component analysis (PCA) for dimensionality reduction and the machine learning classifiers such as SVM, Naive Bayes, Logistic regression and Random forest are used to forecast the winner of the match. The hyper-parameter tuning is performed to highlight the most significant parameters which help in increasing the accuracy of the models used. These models are finally evaluated in terms of accuracy, F1 score, Cohens kappa and AUC. The results show that after hyper parameter tuning Random Forest with F1 score has outperformed all other models with 78% accuracy