EPL Points Prediction Using Machine Learning

<p>This project leverages machine learning techniques to predict match outcomes and the final standings of the English Premier League (EPL). By utilizing historical match data collected through web scraping, we built a robust prediction model that forecasts the results of upcoming fixtures and projects the season's points table.The data pipeline includes comprehensive data preprocessing, feature engineering, and the application of five machine learning algorithms: K-Nearest Neighbors (KNN), Random Forest, AdaBoost, XGBoost, and CatBoost. These models were evaluated using accuracy metrics, achieving up to 96% accuracy in predicting match outcomes. Key features such as team form, home/away advantage, and head-to-head records were engineered to enhance the model's predictive power.The best-performing models, XGBoost and CatBoost, were used to simulate the remaining fixtures, allowing us to generate a predicted points table for the EPL. This project demonstrates the potential of data-driven approaches in sports analytics, offering insights for football clubs, analysts, and enthusiasts looking to understand team performance dynamics.</p>