Research map / Bandits and online learning / Online convex optimization and games
No-Regret Learning in Time-Varying Zero-Sum Games
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| Use the parameter-free two-layer algorithm with DRVU base-learners and stability corrections | Both players in time-varying zero-sum games · Time-varying zero-sum games | individual regret: \widetilde{O}\Big(\sqrt{1+V_T+\min\{P_T,W_T\}}\Big) | Helped |
| Use the parameter-free two-layer algorithm with DRVU base-learners and stability corrections | Both players in time-varying zero-sum games · Time-varying zero-sum games | dynamic NE-regret: \widetilde{O}\big(\min\{\sqrt{(1+V_T)(1+P_T)} + P_T, 1 + W_T\}\big) | Helped |
| Use the parameter-free two-layer algorithm with DRVU base-learners and stability corrections | Both players in time-varying zero-sum games · Time-varying zero-sum games | duality gap: \widetilde{O}\Big(\min\{T^{\frac{3}{4}} \big( 1 +Q_T \big)^{\frac{1}{4}}, T^{\frac{1}{2}}(1+Q_T^{\frac{3}{2}}+P_TQ_T)^{\ | Helped |
| Use the parameter-free two-layer algorithm | A player facing any opponent · Time-varying zero-sum games | individual regret: \widetilde{O}(\sqrt{T}) | Helped |
| Use the parameter-free two-layer algorithm | Both players in a stationary zero-sum game · Stationary zero-sum game | individual regret: \widetilde{O}(1) | Helped |
| Use the parameter-free two-layer algorithm | Both players in a stationary zero-sum game · Stationary zero-sum game | dynamic NE-regret: \widetilde{O}(1) | Helped |
| Use the parameter-free two-layer algorithm | Both players in a stationary zero-sum game · Stationary zero-sum game | duality gap: \widetilde{O}(\sqrt{T}) | Helped |
| Add dummy base-learners that always play a fixed action | The two-layer algorithm · Time-varying zero-sum games | duality gap | Helped |
| Add stability correction terms to the meta-algorithm's losses and predictions | The two-layer algorithm · Time-varying zero-sum games | stability of strategies | Helped |
| Replace NE-regret's minimax value of the cumulative payoff matrix with the sum of each round's minimax value | The performance measure for time-varying games · Two-phase time-varying game | dynamic NE-regret: T/2 | Helped |
Closest papers by meaning
- Competing Against Nash Equilibria in Adversarially Changing Zero-Sum Games (ICML 2019 · similarity 0.92)
- Optimism Without Regularization: Constant Regret in Zero-Sum Games (NeurIPS 2025 · similarity 0.90)
- Convergence of No-Swap-Regret Dynamics in Self-Play (NeurIPS 2024 · similarity 0.89)
- No-regret Learning in Harmonic Games: Extrapolation in the Face of Conflicting Interests (NeurIPS 2024 · similarity 0.89)
- The Harder Path: Last Iterate Convergence for Uncoupled Learning in Zero-Sum Games with Bandit Feedback (ICML 2025 · similarity 0.89)
- Best of Both Worlds: Regret Minimization versus Minimax Play (ICML 2025 · similarity 0.89)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.