Overview / Usage

Welcome to the English Premier League (EPL) Match Result Prediction Project! This project was created as part of my application for the Intel Student Ambassador position. In this project, I have implemented three different classification algorithms - K-Nearest Neighbors (KNN), Naive Bayes, and Decision Trees - to predict the outcomes of EPL matches. The main objective of this project is to explore the effectiveness of these machine learning classifiers in predicting EPL match results. By analyzing historical data and using various features such as team statistics, player performance, and match circumstances, we aim to determine which classifier yields the most accurate predictions.

Methodology / Approach

• Data collection: The data provides information on games from seasons 21/22 and 22/23. The features include ['date', 'time', 'comp', 'round', 'day', 'venue', 'result', 'gf', 'ga','opponent', 'xg', 'xga', 'poss', 'attendance', 'captain', 'formation','referee', 'sh', 'sot', 'dist', 'fk', 'pk', 'pkatt', 'season', 'team'] where 'result' is the target label.
• Preprocessing: Data preprocessing includes cleaning, finding missing values, replacing missing numerical values with median values, categorizing, label encoding, normalizing, and target label balancing using SMOTE for model training.
• Model Implementation: We implement KNN, Naive Bayes, and Decision Tree classifiers.
• Evaluation: Each model's performance is evaluated using accuracy.
• Comparison: We compare the models to determine which one performs the best in predicting EPL match results.
• Documentation: The project is well-documented, including code comments and explanations to help others understand and replicate the work.

Technologies Used

• Python
• Scikit-lean
• Intelex
• Pandas
• Matplotlib
• Seaborn

Repository

https://github.com/Utsabab/Intel-oneAPI-Student-Ambassador