Analysis of team sport matches by machine learning techniques

OData support
Supervisor:
Dr. Toka László
Department of Telecommunications and Media Informatics

Today one of the most popular sport is football. Nothing proves this better than the fact that more than three billion people followed the 2014 FIFA World Football Championship. No wonder clubs are investing huge amounts of money in various technologies in order to become better and better. Over the last decade, teams have realized that they can gain competitive advantage over other teams by using different types of information systems and data-driven models.

In parallel with the spread of Big Data technologies, the complexity and importance of sports analytics has also risen steeply. As a result, revolutionary changes have been made not only in sport as a game, but in the closely related business world such as sports betting. We previously believed that real-time betting is impossible, nowadays it has become commonplace. Gamblers have lots of information on the current form of their favorite team and the latest expert analysis so that they can pinpoint the outcome of a match as accurately as possible. However, this is still a difficult task.

In my thesis, I am using data mining tools to create a data driven predictive model that can predict the outcome of football matches so that it can make a profit. I implemented this model as a web application where users can access the predictions of the model and run simulations before they make their bets.

Downloads

Please sign in to download the files of this thesis.