ICML 2021 accepted papers on Bandits and online learning
88 of the 1,184 papers accepted at ICML 2021 (7.4%) are on Bandits and online learning, compared with 6.4% at ICML 2020. They fall into 4 topics.
See ICML 2021 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 · 27 papers
- Efficient Deviation Types and Learning for Hindsight Rationality in Extensive-Form Games (9 citations)
- Achieving Near Instance-Optimality and Minimax-Optimality in Stochastic and Adversarial Linear Bandits Simultaneously (8 citations)
- On Lower Bounds for Standard and Robust Gaussian Process Bandit Optimization (7 citations)
- Regret and Cumulative Constraint Violation Analysis for Online Convex Optimization with Long Term Constraints (7 citations)
- Memory Efficient Online Meta Learning (6 citations)
- Multi-Receiver Online Bayesian Persuasion (6 citations)
- Beyond $log^2(T)$ regret for decentralized bandits in matching markets (5 citations)
- Online A-Optimal Design and Active Linear Regression (5 citations)
- An Algorithm for Stochastic and Adversarial Bandits with Switching Costs (4 citations)
- A Regret Minimization Approach to Iterative Learning Control (3 citations)
- Non-Exponentially Weighted Aggregation: Regret Bounds for Unbounded Loss Functions (3 citations)
- Robust Learning-Augmented Caching: An Experimental Study (3 citations)
and 15 more in the interactive map.
Multi-armed bandits · 22 papers
- Fairness of Exposure in Stochastic Bandits (9 citations)
- Almost Optimal Anytime Algorithm for Batched Multi-Armed Bandits (8 citations)
- Adapting to Delays and Data in Adversarial Multi-Armed Bandits (7 citations)
- Improved Confidence Bounds for the Linear Logistic Model and Applications to Bandits (4 citations)
- Multi-layered Network Exploration via Random Walks: From Offline Optimization to Online Learning (4 citations)
- Resource Allocation in Multi-armed Bandit Exploration: Overcoming Sublinear Scaling with Adaptive Parallelism (4 citations)
- Robust Pure Exploration in Linear Bandits with Limited Budget (4 citations)
- Combinatorial Blocking Bandits with Stochastic Delays (3 citations)
- Probabilistic Sequential Shrinking: A Best Arm Identification Algorithm for Stochastic Bandits with Corruptions (3 citations)
- Quantile Bandits for Best Arms Identification (3 citations)
- Dichotomous Optimistic Search to Quantify Human Perception (2 citations)
- Dynamic Planning and Learning under Recovering Rewards (2 citations)
and 10 more in the interactive map.
Contextual bandits and Thompson sampling · 20 papers
- Leveraging Good Representations in Linear Contextual Bandits (10 citations)
- MOTS: Minimax Optimal Thompson Sampling (8 citations)
- Optimal Thompson Sampling strategies for support-aware CVaR bandits (7 citations)
- Adversarial Combinatorial Bandits with General Non-linear Reward Functions (6 citations)
- Dynamic Balancing for Model Selection in Bandits and RL (5 citations)
- High-dimensional Experimental Design and Kernel Bandits (5 citations)
- On the Optimality of Batch Policy Optimization Algorithms (5 citations)
- Sparsity-Agnostic Lasso Bandit (5 citations)
- Adapting to misspecification in contextual bandits with offline regression oracles (4 citations)
- Meta-Thompson Sampling (4 citations)
- Top-k eXtreme Contextual Bandits with Arm Hierarchy (4 citations)
- Value-at-Risk Optimization with Gaussian Processes (3 citations)
and 8 more in the interactive map.
RL theory and MDPs · 19 papers
- Provably Efficient Reinforcement Learning for Discounted MDPs with Feature Mapping (23 citations)
- Logarithmic Regret for Reinforcement Learning with Linear Function Approximation (20 citations)
- Model-Free Reinforcement Learning: from Clipped Pseudo-Regret to Sample Complexity (13 citations)
- UCB Momentum Q-learning: Correcting the bias without forgetting (9 citations)
- Adaptive Sampling for Best Policy Identification in Markov Decision Processes (8 citations)
- Near-Optimal Model-Free Reinforcement Learning in Non-Stationary Episodic MDPs (7 citations)
- Near-Optimal Representation Learning for Linear Bandits and Linear RL (6 citations)
- Finding the Stochastic Shortest Path with Low Regret: the Adversarial Cost and Unknown Transition Case (5 citations)
- Risk-Sensitive Reinforcement Learning with Function Approximation: A Debiasing Approach (5 citations)
- Improved Corruption Robust Algorithms for Episodic Reinforcement Learning (4 citations)
- Model-based Reinforcement Learning for Continuous Control with Posterior Sampling (4 citations)
- Randomized Exploration in Reinforcement Learning with General Value Function Approximation (3 citations)
and 7 more in the interactive map.
