NeurIPS 2020 accepted papers on Bandits and online learning
97 of the 1,901 papers accepted at NeurIPS 2020 (5.1%) are on Bandits and online learning, compared with 5.6% at NeurIPS 2019. They fall into 4 topics.
See NeurIPS 2020 in the interactive map
Submitting to NeurIPS? 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 · 50 papers
- Logarithmic Regret Bound in Partially Observable Linear Dynamical Systems (48 citations)
- Information Theoretic Regret Bounds for Online Nonlinear Control (38 citations)
- The Primal-Dual method for Learning Augmented Algorithms (32 citations)
- Learning Strategy-Aware Linear Classifiers (29 citations)
- Online Optimization with Memory and Competitive Control (26 citations)
- Online Bayesian Persuasion (22 citations)
- Optimal Robustness-Consistency Trade-offs for Learning-Augmented Online Algorithms (22 citations)
- Learning Augmented Energy Minimization via Speed Scaling (21 citations)
- Online Influence Maximization under Linear Threshold Model (21 citations)
- Tight last-iterate convergence rates for no-regret learning in multi-player games (16 citations)
- Dynamic Regret of Convex and Smooth Functions (15 citations)
- Bias no more: high-probability data-dependent regret bounds for adversarial bandits and MDPs (13 citations)
and 38 more in the interactive map.
Multi-armed bandits · 20 papers
- An Empirical Process Approach to the Union Bound: Practical Algorithms for Combinatorial and Linear Bandits (24 citations)
- Optimal Best-arm Identification in Linear Bandits (21 citations)
- Adversarial Bandits with Corruptions: Regret Lower Bound and No-regret Algorithm (14 citations)
- Collapsing Bandits and Their Application to Public Health Intervention (12 citations)
- Restless-UCB, an Efficient and Low-complexity Algorithm for Online Restless Bandits (12 citations)
- BanditPAM: Almost Linear Time k-Medoids Clustering via Multi-Armed Bandits (9 citations)
- Adversarial Blocking Bandits (6 citations)
- Crush Optimism with Pessimism: Structured Bandits Beyond Asymptotic Optimality (6 citations)
- Sub-sampling for Efficient Non-Parametric Bandit Exploration (5 citations)
- An Optimal Elimination Algorithm for Learning a Best Arm (4 citations)
- On Regret with Multiple Best Arms (4 citations)
- Batched Coarse Ranking in Multi-Armed Bandits (3 citations)
and 8 more in the interactive map.
Contextual bandits and Thompson sampling · 17 papers
- Model Selection in Contextual Stochastic Bandit Problems (22 citations)
- Adapting to Misspecification in Contextual Bandits (19 citations)
- Latent Bandits Revisited (19 citations)
- Adversarial Attacks on Linear Contextual Bandits (18 citations)
- Inference for Batched Bandits (17 citations)
- High-Dimensional Sparse Linear Bandits (14 citations)
- Locally Differentially Private (Contextual) Bandits Learning (12 citations)
- Stage-wise Conservative Linear Bandits (11 citations)
- An Asymptotically Optimal Primal-Dual Incremental Algorithm for Contextual Linear Bandits (10 citations)
- Efficient Contextual Bandits with Continuous Actions (8 citations)
- Statistical Efficiency of Thompson Sampling for Combinatorial Semi-Bandits (7 citations)
- Online Learning in Contextual Bandits using Gated Linear Networks (6 citations)
and 5 more in the interactive map.
RL theory and MDPs · 10 papers
- Risk-Sensitive Reinforcement Learning: Near-Optimal Risk-Sample Tradeoff in Regret (24 citations)
- Simultaneously Learning Stochastic and Adversarial Episodic MDPs with Known Transition (20 citations)
- Dynamic Regret of Policy Optimization in Non-Stationary Environments (15 citations)
- Upper Confidence Primal-Dual Reinforcement Learning for CMDP with Adversarial Loss (11 citations)
- Adaptive Discretization for Model-Based Reinforcement Learning (9 citations)
- Towards Minimax Optimal Reinforcement Learning in Factored Markov Decision Processes (7 citations)
- Reinforcement Learning in Factored MDPs: Oracle-Efficient Algorithms and Tighter Regret Bounds for the Non-Episodic Setting (6 citations)
- Stateful Posted Pricing with Vanishing Regret via Dynamic Deterministic Markov Decision Processes (5 citations)
- Provably adaptive reinforcement learning in metric spaces (3 citations)
- Almost Optimal Model-Free Reinforcement Learningvia Reference-Advantage Decomposition
