Game Theory and Machine Learning

Despite the success of machine learning, many modern applications now involve data generated or provided by self-interested strategic agents whose objectives depend on the outcome of the learning algorithms. In such cases, standard learning algorithms tend to perform poorly. This is the case for instance in security, where data is generated by attackers who try to evade detection or in applications that involve personal data revealed by privacy-conscious users who try to limit their privacy risks when revealing data. At EURECOM, we work on designing learning algorithms that work better in those cases by using game theory to model the interactions between learning agents and agents generating or providing data and to design algorithms that work optimally given the agents' incentives. Topics include: 

  • Game theory for adversarial classification in security
  • Learning in defense resource allocation games (Blotto)
  • Learning from personal data of privacy-conscious users
  • Statistical learning theory with strategic data

 

Selected publications: 

  • L. Dritsoula, P. Loiseau, and J. Musacchio. A game-theoretic analysis of adversarial classification. arXiv:1610.04972, 2016. [ bib | pdf | http ]

  • V. Kamble, P. Loiseau, and J. Walrand. Regret-optimal strategies for playing repeated games with discounted losses. arXiv:1603.04981, 2016. [ bib | pdf | http ]

  • M. Chessa, J. Grossklags, and P. Loiseau. A game-theoretic study on non-monetary incentives in data analytics projects with privacy implications. In Proceedings of the 28th IEEE Computer Security Foundations Symposium (CSF), 2015. [ bib | pdf | http ]

  • S. Ioannidis and P. Loiseau. Linear regression as a non-cooperative game. In Proceedings of the 9th conference on Web and Internet Economics (WINE), 2013. [ bib | pdf | http ]

  • L. Dritsoula, P. Loiseau, and J. Musacchio. Computing the nash equilibria of intruder classification games. In Proceedings of the third Conference on Decision and Game Theory for Security (GameSec), 2012. [ bib | http ]

Syndicate

Syndicate content

Data Science