Domain: Machine Learning, Web Development, Mobile App Development

Description: This project aimed to build an intelligent decision-support system that provides predictions for football matches based on historical data of the teams involved. The system generates probabilistic predictions for win, loss, or draw outcomes, and allows users to compare their own predictions against those made by the AI model. The solution integrates machine learning algorithms, including XGBoost, Logistic Regression, Naive Bayes, Decision Trees, Random Forests, and Support Vector Machines (SVM). Data scraping was performed using Python to collect match statistics. Initial predictions showed a 55% accuracy rate, which was improved to 80% after enhancing the dataset with additional features and encoding.

The project also includes a web platform and a mobile application:

• Web Platform: Provides predictions for upcoming matches and allows users to make their own predictions.
• Mobile App: Engages users by offering attractive rewards based on their prediction scores, enhancing user motivation and interaction.

Key Features:

• Machine Learning Models: Utilized various algorithms (XGBoost, Logistic Regression, Naive Bayes, Decision Trees, Random Forests, SVM) for accurate predictions.
• Data Collection: Employed web scraping techniques using Python’s BeautifulSoup to gather historical data and match statistics.
• Performance Metrics: Evaluated model performance using metrics such as Confusion Matrix (before and mid-match), Precision, Recall, F1-score, Accuracy, and Balanced Accuracy.
• User Interaction: Users can compare their predictions with AI-generated results and view a detailed summary.
• Improved Prediction Accuracy: Enhanced feature set and data encoding techniques increased prediction accuracy from 55% to 80%.
• Cross-Platform Development: Created both a responsive web platform and a mobile app using Flutter for Android and iOS devices.

Technologies Used:

• Back-End: PHP, Symfony, REST API, MySQL
• Front-End: JavaScript, HTML, CSS, Bootstrap, Twig
• Mobile App: Flutter, Dart
• Machine Learning: BeautifulSoup, Scikit-Learn, Pandas, NumPy, Matplotlib
• Development Environment: Visual Studio Code, Jupyter Notebook

Image App:

pdf : 

Final Dissertation for the Master’s Degree in Computer Science