Year: 2025
Type: article
Abstract: This research project utilized advanced computer algorithms to predict the outcomes of Premier League soccer matches. The dataset containing match data and odds from seasons was processed to handle missing information, select features anzd reduce complexity using Principal Component Analysis. To address imbalances, in the target variable Synthetic Minority Over sampling Technique (SMOTE) was employed. Various machine learning models such as Random Forest, Decision Tree, SVM, XG Boost and Light GBM were evaluated. Model performance was fine tuned using Grid Search CV by adjusting hyperparameters. XG Boost and Light GBM demonstrated the test accuracy at 63% with Random Forest proving dependable. Despite training accuracy rates a decrease in test accuracy indicated overfitting issues. Detailed model performance assessments were provided through confusion matrices and classification reports. The study highlighted the importance of preprocessing and feature engineering alongside model selection for optimal performance improvement. Future steps will focus on enhancing feature engineering techniques exploring methods, regularization strategies and integrating real time data, for research advancements in predicting football match outcomes using machine learning applications.
Author: Agboro Destiny
Language: English
Subfield: Economics and Econometrics
Domain: Social Sciences
Open Access status: green
OpenAlex ID: W4406150324