New machine learning framework optimizes football team performance across

A study in the Journal of Big Data proposes a machine-learning framework to predict football players’ overall performance ratings from dozens of technical, physical, and tactical attributes. Led by Keshav Kaushik of Sharda University with collaborators in India and Italy, the model uses a Multi-Layer Perceptron with two hidden layers. The researchers train the neural network to map player attribute data to a composite rating, drawing its dataset from FIFA 20. They argue the approach treats the game data as a standardized proxy, supported by consistent tracking for more than 17,000 players. In benchmarking against Optimized Linear Regression, LightGBM, Random Forest Regression, and XGBoost, the MLP reached an R² score of 99.13.