Stacked-Based Ensemble Machine Learning Model for Positioning Footballer

• Published in: Arabian journal for science and engineering (2011) Vol. 48; no. 2; pp. 1371 - 1383
• Main Authors: Buyrukoğlu, Selim, Savaş, Serkan
• Format: Journal Article
• Language: English
• Published: Berlin/Heidelberg Springer Berlin Heidelberg 01.02.2023; Springer Nature B.V
• Subjects: Algorithms; Artificial neural networks; Chi-square test; Engineering; Feature selection; Humanities and Social Sciences; Machine learning; multidisciplinary; Research Article-Computer Engineering and Computer Science; Science; Feature selection; Meta-learner; Chi-square; Ensemble learning; Footballer position classification
• ISSN: 2193-567X; 1319-8025; 2191-4281
• DOI: 10.1007/s13369-022-06857-8

Due to the high performance of machine learning (ML) methods in different disciplines, these methods have been used frequently in various sports fields, especially in the last decade. Researchers have used ML algorithms in football on various subjects such as match result prediction, estimation of factors affecting match results, prediction of league standings, and analysis of the performances of football players. However, there has not been enough work on determining the position of the football player, which is one of the leading problems for coaches in football. Therefore, this study aims to classify footballer positions employing a stacked ensemble ML model using the FIFA’19 game dataset. To achieve this aim, a two-stage application is followed. In the first stage, 10 features are selected using four different feature selection algorithms. In the second stage, Deep Neural Networks, Random Forest, and Gradient Boosting were used as single-based algorithms, and Logistic Regression was employed as a meta-learner in the stacked model. The results show that the combination of the Chi-square feature selection technique and the stacked-based ensemble learning model yielded the best accuracy (83.9%). The findings emphasize the validity and robustness of our stacked ensemble learning model to determine the positions of footballers.