CS 364 A : Algorithmic Game Theory
نویسنده
چکیده
Last lecture we proved that coarse correlated equilibria (CCE) are tractable, in a satisfying sense: there are simple and computationally efficient learning procedures that converge quickly to the set of CCE. Of course, if anything in our equilibrium hierarchy (Figure 1) was going to be tractable, it was going to be CCE, the biggest set. The good researcher is never satisfied and always seeks stronger results. What can we say if we zoom in to the next-biggest set, the correlated equilibria? The first part of this lecture shows that correlated equilibria are also tractable. We’ll give computationally efficient — if not quite as simple — learning procedures that converge fairly quickly to this set.
منابع مشابه
CS 364 A : Algorithmic Game Theory Lecture # 17 : No - Regret Dynamics ∗
This lecture continues to study the questions introduced last time. Do strategic players reach an equilibrium of a game? How quickly? By what learning processes? Positive results on these questions justify equilibrium analysis, including bounds on the price of anarchy. Last lecture focused on best-response dynamics. These dynamics are most relevant for potential games, which cover many but not ...
متن کاملCS 364 A : Algorithmic Game Theory Lecture # 18 : From External Regret to Swap Regret and the Minimax Theorem ∗
Last lecture we proved that coarse correlated equilibria (CCE) are tractable, in a satisfying sense: there are simple and computationally efficient learning procedures that converge quickly to the set of CCE. Of course, if anything in our equilibrium hierarchy (Figure 1) was going to be tractable, it was going to be CCE, the biggest set. The good researcher is never satisfied and always seeks s...
متن کاملCS 364 A : Algorithmic Game Theory Lecture # 13 : Potential Games ; A Hierarchy of Equilibria ∗
Last lecture we proved that every pure Nash equilibrium of an atomic selfish routing game with affine cost functions (of the form ce(x) = aex + be with ae, be ≥ 0) has cost at most 5 2 times that of an optimal outcome, and that this bound is tight in the worst case. There can be multiple pure Nash equilibria in such a game, and this bound of 5 2 applies to all of them. But how do we know that t...
متن کاملCS 364 A : Algorithmic Game Theory Lecture # 6 : Simple Near - Optimal Auctions ∗
for each input v. If every Fi is regular, meaning that the corresponding virtual valuation function φi(vi) = vi − 1−Fi(vi) fi(vi) is strictly increasing, then the virtual welfare-maximizing allocation rule is monotone and, after we define suitable payments, maximizes expected revenue over all DSIC auctions. This characterization of optimal auctions can be extended to irregular distributions, bu...
متن کاملCS 364 A : Algorithmic Game Theory Lecture # 16 : Best - Response Dynamics ∗
Affirmative answers to these questions are important because they justify equilibrium analysis. Properties of equilibria, such as a near-optimal objective function value, are not obviously relevant when players fail to find one. More generally, proving that natural learning algorithms converge quickly to an equilibrium lends plausibility to the predictive power of an equilibrium concept. To rea...
متن کامل