Jesse Davis


Department of Computer Science, KU Leuven


Jesse Davis is an associate professor in the Department of Computer Science at KU Leuven in Belgium. He received his bachelor’s degree from Williams College (USA) and his masters’ and PhD from the University of Wisconsin – Madison (USA). He completed a three year post doc at the University of Washington (USA). He is also a co-founder of Activ84Health, an awarding-winning startup whose mission is to motivate elderly individuals to remain physically active.


Jesse’s main research interests are in machine learning, artificial intelligence, and data mining. Concretely, his research focuses on (1) learning from and mining uncertain, highly structured data, (2) transfer learning, (3) learning from positive and unlabeled data, and (4) anomaly detection. Furthermore, he is interested in applying techniques from these fields to address problems in sports, health, and their intersection. Here, he focuses on problems such as valuing the actions performed by players during a game, discovering common tactics employed by teams, and analyzing sensor data (e.g., accelerometer, heartrate, GPS) collected from athletes during training sessions or competitions.


  1. Tom Decroos, Jan Van Haaren and Jesse Davis (2018). Automatic Discovery of Tactics in Spatio-Temporal Soccer Match Data. To appear in the Proceedings of the 24th ACM Conference on on Knowledge Discovery and Data Mining (KDD 2018).
  2. Tim Op De Beeck, Wannes Meert, Kurt Schutte, Benedicte Vanwanseele and Jesse Davis (2018). Fatigue Prediction in Outdoor Runners Via Machine Learning and Sensor Fusion. In the Proceedings of the 24th ACM Conference on on Knowledge Discovery and Data Mining (KDD 2018).
  3. Arne Jaspers, Tim Op De Beéck, Michel S. Brink, Wouter G. P. Frencken, Filip Staes, Jesse Davis, and Werner Helsen (2018). Relationships between the External and Internal Training Load in Professional Soccer: What Can We Learn from Machine Learning? International Journal of Sports Physiology and Performance.
  4. Tom Decroos, Vladimir Dzyuba, Jan Van Haaren, and Jesse Davis (2017). Predicting soccer highlights from spatio-temporal match event streams. In the Proceedings of the Thirty-First AAAI Conference on Artificial Intelligence (AAAI 2017).