Curriculum
Computer Science for Societal Challenges and Innovation, XXXIX series
Grant sponsor
Università degli Studi di Padova
Supervisor

GianAntonio Susto
Co-supervisor
s
Alberto Testolin
Contact
alberto.sinigaglia@studenti.unipd.it

Project description
Deep Reinforcement Learning has been recently very successful in solving tasks thought to be decades far from being tackled. However, when considering multiple actors collaborating or competing, the formulation of the problem becomes significantly more complicated, as the optimal solution is known to be computationally hard in the general case, and thus special techniques need to be developed to help convergence, especially when considered the conjunction with Deep Learning.