Rating Prediction of Football Players using Machine Learning

This research Analysis the prediction of football player ratings through the application of diverse machine learning algorithms. Rating systems for sports teams have gathered considerable attention in academic research. The approach used by the authors of this paper serves as an effort to streamline scouts and performance analytics. Leveraging linear regression, decision tree regressor, random forest regressor, gradient boosting regressor, support vector regressor, voting regressor, ridge regression, lasso regression, k-nearest neighbours' regression, Huber regression and elastic-net regression. The Analysis explores the efficiency of each algorithm and concludes that Support Vector Regressor algorithm performs the best with 91.84% accuracy on the testing data followed by the Gradient Boosting Regressor with 90.78%, Voting Regressor with 91.68% and Random Forest Regressor with 88.89%. Apart from them the K-Nearest Neighbours Regression Algorithm highly overfits the model with 100% accuracy on the training set and 70.71%. The conclusions drawn underscore the critical importance of judiciously selecting algorithms tailored to the specific characteristics of the dataset for precise and reliable player rating predictions.