We consider the celebrated Blackwell Approachability Theorem for two-player games with vector payoffs. We show that Blackwell's result is equivalent, via efficient reductions, to the existence of "no-regret" algorithms for Online Linear Optimization.
Nov 8, 2010
In the present paper we prove, to the contrary, that Blackwell's Approachability Theorem is equivalent, in a very strong sense, to no-regret learning for the ...
Oct 22, 2024 · We show that this relationship is in fact much stronger, that Blackwell's result is equivalent to, in a very strong sense, the problem of regret ...
We show that this relationship is in fact much stronger, that Blackwell's result is equivalent to, in a very strong sense, the problem of regret minimization.
Missing: Low- | Show results with:Low-
Nov 8, 2010 · We show that Blackwell's result is equivalent, via efficient reductions, to the existence of “no- regret” algorithms for Online Linear ...
Aug 13, 2018 · Jacob D. Abernethy, Peter L. Bartlett, Elad Hazan: Blackwell Approachability and Low-Regret Learning are Equivalent.
These algorithms rely on Blackwell's primal condition – existence of a separating hyperplane – and essentially require computing a projection direction onto S.
Feb 17, 2023 · Blackwell's approachbility theorem is important because it has very deep connections to online learning and regret minimization. As Vohra states ...
so if we can make the average payoff in the approachability problem approach the set S, we get low regret in the online learning problem. To apply Blackwell's ...
Blackwell approachability and low-regret learning are equivalent. In Conference on Learning Theory (COLT), pages 27–46,. June 2011. R.J. Aumann and M ...