Coşkun KUŞ, İsmail Hakkı KINALIOĞLU

Prediction of UEFA champions league elimination rounds winners using machine learning algorithms

In this study, the teams that qualified for the next round as a result of two-legged matchupsare predicted using the data collected from the UEFA (Union of European FootballAssociations) Champions League group stage matches. The study contributes to the literaturein terms of variety of methods used and content of the dataset compared to other studiesconducted on football data. It is also a pioneering study to predict the outcome of a two-leggedmatchup. The data are collected from the matches played in the Champions Leagueorganizations held between 2010-2018. Classification methods as Artificial Neural Network,K-Nearest Neighbors, Logistic Regression Analysis, Naive Bayes Classifier, Random Forestand Support Vector Machine are used for the prediction. Two applications are carried out totest the successes of the classification models. In the first application, the most successfulmethod is naive bayes classifier (86.66%) and in the second application, the most successfulmethod is random forest (74.81%).

PDF

___

[1] Caraffa, A., Cerulli G., Projetti M., Aisa G. and Rizzo, A., Prevention of anterior cruciate ligament injuries in soccer, Knee Surgery, Sports Traumatology, Arthroscopy, 4 (1) (1996) 19-21.
[2] Heidt R. S., Sweeterman L. M., Carlonas R. L., Traub J. A. and Tekulve F. X., Avoidance of Soccer Injuries with Preseason Conditioning, The American Journal of Sports Medicine, 28 (5) (2000) 659-662.
[3] Houston R. G., Wilson D. P., Income, leisure and proficiency: an economic study of football performance, Applied Economics Letters, 9 (14) (2002) 939-943.
[4] Junge A., Dvorak J., Soccer Injuries, Sports Medicine, 34 (13) (2004) 929-938.
[5] Decrop A., Derbaix C., Pride in contemporary sport consumption: a marketing perspective, Journal of the Academy of Marketing Science, 38 (5) (2010) 586-603.
[6] Goddard J., Regression models for forecasting goals and match results in association football, International Journal of forecasting, 21 (2) (2005) 331-340.
[7] Huang K.-Y., Chen K.-J., Multilayer perceptron for prediction of 2006 world cup football game, Advances in Artificial Neural Systems, 2011 (2011) 11.
[8] Timmaraju A. S., Palnitkar A. and Khanna V., Game ON! Predicting English Premier League Match Outcomes, (2013).
[9] Ulmer B., Fernandez M. and Peterson M., Predicting Soccer Match Results in the English Premier League, Doctoral dissertation, Ph. D. dissertation, Stanford, (2013).
[10] Lata K., Gupta P., Forecasting English Premier League Match Results, International Journal of Modern Trends in Engineering and Research, 03 (02) (2016) 261-268.
[11] Igiri C. P., Support Vector Machine—Based Prediction System for a Football Match Result, IOSR Journal of Computer Engineering, 17 (3) (2015) 21-26.
[12] Baboota R., Kaur H., Predictive analysis and modelling football results using machine learning approach for English Premier League, International Journal of Forecasting, 35 (2) (2019) 741-755.
[13] Rahman M. A., A deep learning framework for football match prediction, SN Applied Sciences, 2 (2) (2020) 165.
[14] Rotshtein A. P., Posner M., Rakityanskaya A., Football predictions based on a fuzzy model with genetic and neural tuning, Cybernetics and Systems Analysis, 41 (4) (2005) 619-630.
[15] Joseph A., Fenton N. E. and Neil M., Predicting football results using Bayesian nets and other machine learning techniques, Knowl-Based Syst, 19 (7) (2006) 544-553.
[16] Hvattum L. M., Arntzen H., Using ELO ratings for match result prediction in association football, International Journal of forecasting, 26 (3) (2010) 460-470.
[17] Owen A., Dynamic bayesian forecasting models of football match outcomes with estimation of the evolution variance parameter, IMA Journal of Management Mathematics, 22 (2) (2011) 99-113.
[18] Constantinou A. C., Fenton N. E. and Neil M., pifootball: A Bayesian network model for forecasting Association Football match outcomes, Knowl-Based Syst, 36 (2012) 322-339.
[19] Igiri C. P., Nwachukwu E. O., An improved prediction system for football a match result, IOSR Journal of Engineering, 4 (2014) 12-20.
[20] Koopman S. J., Lit R. A., dynamic bivariate Poisson model for analysing and forecasting match results in the English Premier League, Journal of the Royal Statistical Society: Series A, 178 (1) (2015) 167-186.
[21] Amadin F., Obi, J. C.i English Premier League (EPL) Soccer Matches Prediction using An Adaptive Neuro-Fuzzy Inference System (ANFIS), Transactions on Machine Learning and Artificial Intelligence, 3 (2) (2015) 34.
[22] Robertson S., Back N. and Bartlett J. D., Explaining match outcome in elite Australian Rules football using team performance indicators, Journal of Sports Sciences, 34 (7) (2016) 637- 644.
[23] Gevaria K., Sanghavi H., Vaidya S. and Deulkar K., Football Match Winner Prediction. (2015).
[24] Prasetio D., Dra. Harlili M. S., in Advanced Informatics: Concepts, Theory And Application, International Conference On., 2016, 1-5.
[25] Martins R. G. et al., Exploring polynomial classifier to predict match results in football championships, Expert Syst. Appl., 83 (2017) 79- 93.
[26] Bunker R. P., Thabtah F., A machine learning framework for sport result prediction, Applied Computing and Informatics, 15 (1) (2017) 27-33.
[27] Mennis J., Guo D., Spatial data mining and geographic knowledge discovery—An introduction, Computers, Environment and Urban Systems, 33 (6) (2009) 403-408.
[28] Aggarwal C. C., Data classification, algorithms and applications, New York: CRC Press, 2014.
[29] Elmas Ç., Yapay zeka uygulamaları: (yapay sinir ağı, bulanık mantık, genetik algoritma), İstanbul: Seçkin Yayıncılık, 2010.
[30] Haykin S., Neural networks: a comprehensive foundation, Prentice Hall PTR, 1994.
[31] Peterson L. E., K-nearest neighbor. Scholarpedia, 4 (2) (2009) 1883.
[32] Fix E., Hodges Jr, J. L., Discriminatory analysisnonparametric discrimination: consistency properties, California Univ Berkeley, 1951.
[33] Cover T. M., Hart P. E., Nearest neighbor pattern classification, IEEE transactions on information theory, 13 (1) (1967) 21-27.
[34] Song Y., Huang J., Zhou D., Zhan H. and Giles C. L., in European Conference on Principles of Data Mining and Knowledge Discovery, 248-264.
[35] Cramer J. S., in Tinbergen Institute Discussion Paper, No. 02-119/4, Tinbergen Institute, Amsterdam and Rotterdam, 2002.
[36] Hosmer D. W., Lemeshow S., Applied Logistic Regression, New York: John Wiley & Sons, 2000.
[37] Özdamar K., Paket programlar ile istatistiksel veri analizi (çok değişkenli analizler), Kaan Kitabevi, 2004).
[38] Zhang H. The optimality of naive Bayes, AA 1 (2) (2004) 3.
[39] Kantarcıoglu M., Vaidya J. and Clifton C. in IEEE ICDM workshop on privacy preserving data mining, 3-9.
[40] Rish I., in IJCAI 2001 Workshop on Empirical Methods in Artificial Intelligence. 41-46.
[41] Ho T. K, in Proceedings of 3rd international conference on document analysis and recognition. 278-282 (IEEE).
[42] Breiman L., Random Forests, UC Berkeley TR567 (1999).
[43] Breiman L., Random Forests, Mach Learn 45 (1) (2001) 5-32.
[44] Schölkopf B., Smola A. J. and Bach F., Learning with kernels: support vector machines, regularization, optimization, and beyond. MIT press, 2002.
[45] Kavzoğlu T., Çölkesen İ., Destek vektör makineleri ile uydu görüntülerinin sınıflandırılmasında kernel fonksiyonlarının etkilerinin incelenmesi, Harita Dergisi 144 (7) (2010) 73-82.
[46] Yang Z.-X., Shao Y.-H. and Zhang X.-S., Multiple birth support vector machine for multiclass classification, Neural Computing and Applications, 22 (1) (2013) 153-161.
[47] Crammer K. and Singer Y., On the algorithmic implementation of multiclass kernel-based vector machines, Journal of machine learning research ,2 (Dec) (2001) 265-292.
[48] Platt J. C., Cristianini N. and Shawe-Taylor J., in Advances in neural information processing systems. 547-553.
[49] Dietterich T. G., Bakiri G., Solving multiclass learning problems via error-correcting output codes, Journal of artificial intelligence research, 2 (1994) 263-286.
[50] Crammer K., Singer Y., On the learnability and design of output codes for multiclass problems, Mach Learn., 47 (2-3) (2002) 201-233.
[51] Allwein E. L., Schapire R. E. and Singer Y., Reducing multiclass to binary: A unifying approach for margin classifiers, Journal of machine learning research, 1 (2000) 113-141.
[52] Özkan Y., Veri madenciliği yöntemleri, Papatya Yayıncılık Eğitim, 2016.
[53] Hall M. A. Correlation-based feature selection for machine learning. (1999).
[54] DeLong, E. R., DeLong D. M. and ClarkePearson D. L., Comparing the areas under two or more correlated receiver operating characteristic curves: a nonparametric approach, Biometrics, 44 (3) (1988) 837-845