ICML 2022 accepted papers on Bandits and online learning
76 of the 1,233 papers accepted at ICML 2022 (6.2%) are on Bandits and online learning, compared with 7.4% at ICML 2021. They fall into 4 topics.
See ICML 2022 in the interactive map
Submitting to ICML? Checklist →
Working on something in this area? Describe your idea in scime atlas to see the closest papers by meaning and how crowded the spot has become.
Topics and papers
Online convex optimization and games · 24 papers
- Online Learning for Min Sum Set Cover and Pandora’s Box (5 citations)
- Faster Fundamental Graph Algorithms via Learned Predictions (4 citations)
- Near-Optimal Learning of Extensive-Form Games with Imperfect Information (3 citations)
- Equivalence Analysis between Counterfactual Regret Minimization and Online Mirror Descent (2 citations)
- No-Regret Learning in Time-Varying Zero-Sum Games (2 citations)
- Online Algorithms with Multiple Predictions (2 citations)
- Parsimonious Learning-Augmented Caching (2 citations)
- A Marriage between Adversarial Team Games and 2-player Games: Enabling Abstractions, No-regret Learning, and Subgame Solving (1 citation)
- Learning Markov Games with Adversarial Opponents: Efficient Algorithms and Fundamental Limits (1 citation)
- Online Balanced Experimental Design (1 citation)
- Regret Minimization with Performative Feedback (1 citation)
- Risk-Averse No-Regret Learning in Online Convex Games (1 citation)
and 12 more in the interactive map.
Contextual bandits and Thompson sampling · 20 papers
- A Simple Unified Framework for High Dimensional Bandit Problems (5 citations)
- Thresholded Lasso Bandit (5 citations)
- Langevin Monte Carlo for Contextual Bandits (4 citations)
- Contextual Bandits with Smooth Regret: Efficient Learning in Continuous Action Spaces (2 citations)
- Linear Bandit Algorithms with Sublinear Time Complexity (2 citations)
- Contextual Information-Directed Sampling (1 citation)
- Feature and Parameter Selection in Stochastic Linear Bandits (1 citation)
- Inverse Contextual Bandits: Learning How Behavior Evolves over Time (1 citation)
- Universal and data-adaptive algorithms for model selection in linear contextual bandits (1 citation)
- Breaking the $\sqrtT$ Barrier: Instance-Independent Logarithmic Regret in Stochastic Contextual Linear Bandits
- Contextual Bandits with Large Action Spaces: Made Practical
- Coordinated Attacks against Contextual Bandits: Fundamental Limits and Defense Mechanisms
and 8 more in the interactive map.
Multi-armed bandits · 19 papers
- On the Finite-Time Performance of the Knowledge Gradient Algorithm (3 citations)
- Strategies for Safe Multi-Armed Bandits with Logarithmic Regret and Risk (3 citations)
- When Are Linear Stochastic Bandits Attackable? (2 citations)
- Adaptive Best-of-Both-Worlds Algorithm for Heavy-Tailed Multi-Armed Bandits (1 citation)
- Batched Dueling Bandits (1 citation)
- Interactively Learning Preference Constraints in Linear Bandits (1 citation)
- Multiple-Play Stochastic Bandits with Shareable Finite-Capacity Arms (1 citation)
- Near-Optimal Algorithms for Autonomous Exploration and Multi-Goal Stochastic Shortest Path (1 citation)
- Stochastic Rising Bandits (1 citation)
- Thompson Sampling for (Combinatorial) Pure Exploration (1 citation)
- Congested Bandits: Optimal Routing via Short-term Resets
- Nested Bandits
and 7 more in the interactive map.
RL theory and MDPs · 13 papers
- Regret Bounds for Stochastic Shortest Path Problems with Linear Function Approximation (4 citations)
- First-Order Regret in Reinforcement Learning with Linear Function Approximation: A Robust Estimation Approach (3 citations)
- From Dirichlet to Rubin: Optimistic Exploration in RL without Bonuses (3 citations)
- Learning Infinite-horizon Average-reward Markov Decision Process with Constraints (3 citations)
- Learning Stochastic Shortest Path with Linear Function Approximation (2 citations)
- Nearly Minimax Optimal Reinforcement Learning with Linear Function Approximation (2 citations)
- Sample-Efficient Reinforcement Learning with loglog(T) Switching Cost (2 citations)
- Cooperative Online Learning in Stochastic and Adversarial MDPs
- Dynamic Regret of Online Markov Decision Processes
- Improved No-Regret Algorithms for Stochastic Shortest Path with Linear MDP
- Nearly Optimal Policy Optimization with Stable at Any Time Guarantee
- Stabilizing Q-learning with Linear Architectures for Provable Efficient Learning
and 1 more in the interactive map.
