Dario Bauso

Engineering and Technology Institute, University of Groningen, The Netherlands

Talk: Gradient-Based Learning Dynamics in Coalitional Games with Transferable Utility

We study a group of players making joint decisions under uncertainty and learning via gradient dynamics how to reach the maximal profit and the revenue gets allocated to the players while the game is running. In the first scenario, uncertainty is modeled by a known distribution, allowing players to update decisions via deterministic gradient ascent dynamics. We show that the learning dynamics and value function converge linearly to the optimum. In the second scenario, only samples are available, and players use stochastic gradient ascent dynamics. We prove that the stochastic objective, evaluated at the average of the stochastic gradient ascent dynamics, converges sublinearly to the optimum. We further propose a fair profit allocation mechanism, based on the Shapley value, and show that this learning-allocation mechanism converges to its optimum in both the deterministic and stochastic cases. Numerical results support our theoretical findings.

Bio: Dario Bauso has received the Laurea degree in Aeronautical Engineering in 2000 and the Ph.D. degree in Automatic Control and System Theory in 2004 from the University of Palermo, Italy. Since 2018 he has been with the Jan C. Willems Center for Systems and Control, ENTEG, Faculty of Science and Engineering, University of Groningen (The Netherlands), where he is currently Full Professor and Chair of Operations Research for Engineering Systems. Since 2005 he has also been with the Dipartimento di Ingegneria, University of Palermo (Italy). Since 2018 he has been a guest professor at Keio University, Japan. His research interests are in the field of Optimization, Optimal and Distributed Control, and Game Theory. Bauso was an Associate Editor of IEEE Transactions on Automatic Control from 2011 to 2016, of  IFAC Automatica from 2015 to 2021, of IEEE Control Systems Letters from 2016 to 2021, of Dynamic Games and Applications from 2011 to 2022, and is Associate Editor of Journal of Dynamics and Games since 2019.