**Date: **
17:00,
Thursday, March 10, 2016

**Speaker: **
Guillermo A. Pérez

**Venue: **IST Austria

In this talk we consider the problem of synthesising strategies which ensure minimal regret for a player in a Discounted-Sum Game. Such strategies minimise the difference between the payoff obtained by the player and the payoff he could have achieved if he had known the strategy of his adversary in advance. We give algorithms for the general problem of computing the minimal regret of a player as well as several variants depending on which strategies the adversary is permitted to use. We also consider the special case of synthesising regret-free strategies in the same scenarios.