Articles In Press / Online First                   Back to the articles list | Back to browse issues page

XML Print


1- Multi disciplinary laboratory of education science and training engineering , mohamed.rebbouj@enscasa.ma
2- Multi disciplinary laboratory of education science and training engineering
Abstract:   (1204 Views)
Background. Sport results predictive analysis is based on betting apps outcomes and has not yet been examined academically by concerned organizations in Morocco. Objectives. This study aims to predict a football national league ranking using Machine Learning regression model with Elastic Net algorithm, where we determine the important features’ weight on prediction. Methods. A dataset of historical scores of 8 standing teams since the 2009/2010 season was manually filled in and categorized into 9 columns: season, team, points, goal difference (+/-), matches played (M), matches won (W), matches drawn (D), matches lost (L), goals for (F) and goals against (A). Then preprocessed into Categorical data, categorical Hash and numeric. Results. the machine learning analysis results in R2 score = 0.999, NRMSE= 0.001 and Spearman correlation = 0.997. However, the predicted ranking was correct about 5 from 8 compared to the actual results. Conclusion. The Ranking prediction has been accurate by 75% in actual results compared to the regression analysis outcomes. Which proves the quality of data needs to be more precise by including other parameters.
Keywords: Football Ranking, Machine Learning, Regression, Prediction

This article is"Accepted Uncorrected Proofs"

Full-Text [PDF 1365 kb]   (571 Downloads)    
Type of Study: Original Article | Subject: Sport Management and its related branches
Received: 2023/06/21 | Accepted: 2023/01/2

Send email to the article author


Rights and permissions
Creative Commons License This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.

© 2024 CC BY-NC 4.0 | Annals of Applied Sport Science

Designed & Developed by : Yektaweb