Predicting football match outcomes with fantasy league data and deep learning

Predicting and betting on the outcome of football matches has been around as long as the game itself. However, with the explosion of online betting over the past 15 years, the stakes are much higher, and the person who accurately predicts the outcome has a chance of making sizable profits. With data from football getting better each year and as more data is gathered, there was only a matter of time until people would start exploring the use of machine learning for predicting match results. Here we present such a method. Using data from the most popular fantasy game in the world, Fantasy Premier League, combined with already existing data currently used to predict the outcome of matches. We use a Recurrent Neural Network (RNN) with Gated Recurrent Unit (GRU) layers to predict the outcome of matches. Our combined method shows good results, improving significantly on state of the art.

Author dc:creator

• Sveinn Henrik Kristinsson 1993-

Contributors dc:contributor

• Háskólinn í Reykjavík

• Tölvunarfræði
• Meistaraprófsritgerðir
• Vélrænt nám
• Spálíkön
• Knattspyrna
• Computer science
• Machine learning
• Forecasting
• Modeling
• Football

Language dc:language.iso

en

Handle dc:identifier.uri

http://hdl.handle.net/1946/43314

OAI identifier oai:identifier

oai:skemman.is:1946/43314