DeepAI AI Chat
Log In Sign Up

Conic Blackwell Algorithm: Parameter-Free Convex-Concave Saddle-Point Solving

05/27/2021
by   Julien Grand-Clément, et al.
0

We develop new parameter and scale-free algorithms for solving convex-concave saddle-point problems. Our results are based on a new simple regret minimizer, the Conic Blackwell Algorithm^+ (CBA^+), which attains O(1/√(T)) average regret. Intuitively, our approach generalizes to other decision sets of interest ideas from the Counterfactual Regret minimization (CFR^+) algorithm, which has very strong practical performance for solving sequential games on simplexes. We show how to implement CBA^+ for the simplex, ℓ_p norm balls, and ellipsoidal confidence regions in the simplex, and we present numerical experiments for solving matrix games and distributionally robust optimization problems. Our empirical results show that CBA^+ is a simple algorithm that outperforms state-of-the-art methods on synthetic data and real data instances, without the need for any choice of step sizes or other algorithmic parameters.

READ FULL TEXT

page 1

page 2

page 3

page 4

02/24/2022

Solving optimization problems with Blackwell approachability

We introduce the Conic Blackwell Algorithm^+ (CBA^+) regret minimizer, a...
09/10/2018

Online Convex Optimization for Sequential Decision Processes and Extensive-Form Games

Regret minimization is a powerful tool for solving large-scale extensive...
02/19/2020

Stochastic Regret Minimization in Extensive-Form Games

Monte-Carlo counterfactual regret minimization (MCCFR) is the state-of-t...
10/11/2021

Equivalence Analysis between Counterfactual Regret Minimization and Online Mirror Descent

Counterfactual Regret Minimization (CFR) is a kind of regret minimizatio...
11/06/2018

Regret Circuits: Composability of Regret Minimizers

Regret minimization is a powerful tool for solving large-scale problems;...
11/06/2018

Composability of Regret Minimizers

Regret minimization is a powerful tool for solving large-scale problems;...
10/07/2019

Combining No-regret and Q-learning

Counterfactual Regret Minimization (CFR) has found success in settings l...