PREDICTION OF FOOTBALL PLAYER’S POSITION USING MACHINE LEARNING ALGORITHMS

Abstract

Football is a collective sport that is played between two teams, composed of eleven players each. Although players play in a predetermined position, they can quickly move to another position. In this paper, the prediction of the best position for a football player was made based on his physical and mental characteristics. This approach's primary motivation is to provide additional information to coaches whose clubs are faced with many injured players. Each player was represented with a vector of 65 attributes, and the following models: Multinomial Logistic Regression, K-Nearest Neighbors, Random Forest, Gaussian Naive Bayes, Support Vector Machine were experimented with.