Call for Papers: "Machine Learning for Soccer"

Call for Papers for a Special Issue in the Machine Learning Journal: "Machine Learning for Soccer".
Special issue editors: Werner Dubitzky, Daniel Berrar, Philippe Lopes, Jesse Davis.

Call

The Machine Learning journal invites submissions of original contributions to machine learning research for soccer analytics. Soccer1 is the biggest global sport and is a fast-growing multibillion dollar industry. The annual revenue of European football clubs alone is estimated at $27bn. Data science and analytics are being more frequently employed on both the club and national levels to improve performance, equipment, marketing, scouting, etc. In conjunction with this special issue, we will offer a machine learning challenge task where the goal is to predict the outcomes of future matches based on a data set of over 200,000 soccer matches from soccer leagues around world. This special issue solicits papers about machine learning approaches for all aspects of soccer, including (but not limited to) topics such as: 

  • Predicting the outcome of individual soccer matches, entire competitions and tournaments;
  • Predicting the performance of individual soccer players, entire teams, and team elements;
  • Aiding the design, planning and selection of competitive soccer strategies, tactics and teams;
  • Developing and improving the performance of soccer players and teams;
  • Evaluating and profiling young talented soccer players (scouting);
  • Improving the development of young soccer players and integration of new players;
  • Analyzing complex data (video, sensors, texts, etc.) from soccer players and matches;
  • Designing soccer training programs that help avoid injuries and improve the recovery of injured players;
  • Predicting the outcome of future soccer matches as part of this special issue’s prediction challenge (see text below). 

Prediction Challenge

This special issue includes a machine learning research challenge task based on a training data set of over 200,000 soccer matches. Challenge participants should use this data set to construct a model that predicts the outcome of a defined set of future soccer matches. This challenge presents a unique real-world machine learning prediction problem and it involves solving various machine learning tasks: data integration/fusion, feature modeling/learning, and outcome prediction. People interested in participating in the challenge should contact the guest editors immediately to receive the training data set and a description of the challenge. The challenge participants will receive an updated training set and a prediction data set on 22 March 2017 (which includes the fixtures of future matches with unknown outcome) and are required to submit their predictions by midnight on 30 March 2017 Central European Time (CET). Details about the data and evaluation criteria will be provided upon expression of interest (see Important Dates below). 

 

More information in the full call.

Group: