Timezone: »
Motivated by cognitive radio networks, we consider the stochastic multiplayer multi-armed bandit problem, where several players pull arms simultaneously and collisions occur if one of them is pulled by several players at the same stage. We present a decentralized algorithm that achieves the same performance as a centralized one, contradicting the existing lower bounds for that problem. This is possible by ``hacking'' the standard model by constructing a communication protocol between players that deliberately enforces collisions, allowing them to share their information at a negligible cost. This motivates the introduction of a more appropriate dynamic setting without sensing, where similar communication protocols are no longer possible. However, we show that the logarithmic growth of the regret is still achievable for this model with a new algorithm.
Author Information
Etienne Boursier (ENS Paris Saclay)
Vianney Perchet (ENSAE & Criteo AI Lab)
Related Events (a corresponding poster, oral, or spotlight)
-
2019 Spotlight: SIC-MMAB: Synchronisation Involves Communication in Multiplayer Multi-Armed Bandits »
Thu. Dec 12th 12:10 -- 12:15 AM Room West Exhibition Hall A
More from the Same Authors
-
2021 Spotlight: Online Sign Identification: Minimization of the Number of Errors in Thresholding Bandits »
Reda Ouhamma · Odalric-Ambrym Maillard · Vianney Perchet -
2021 Spotlight: Decentralized Learning in Online Queuing Systems »
Flore Sentenac · Etienne Boursier · Vianney Perchet -
2022 Poster: Gradient flow dynamics of shallow ReLU networks for square loss and orthogonal inputs »
Etienne Boursier · Loucas PILLAUD-VIVIEN · Nicolas Flammarion -
2022 Poster: Active Labeling: Streaming Stochastic Gradients »
Vivien Cabannes · Francis Bach · Vianney Perchet · Alessandro Rudi -
2021 Poster: Local Differential Privacy for Regret Minimization in Reinforcement Learning »
Evrard Garcelon · Vianney Perchet · Ciara Pike-Burke · Matteo Pirotta -
2021 Poster: ROI Maximization in Stochastic Online Decision-Making »
Nicolò Cesa-Bianchi · Tom Cesari · Yishay Mansour · Vianney Perchet -
2021 Poster: Making the most of your day: online learning for optimal allocation of time »
Etienne Boursier · Tristan Garrec · Vianney Perchet · Marco Scarsini -
2021 Poster: Stochastic Online Linear Regression: the Forward Algorithm to Replace Ridge »
Reda Ouhamma · Odalric-Ambrym Maillard · Vianney Perchet -
2021 Poster: Online Sign Identification: Minimization of the Number of Errors in Thresholding Bandits »
Reda Ouhamma · Odalric-Ambrym Maillard · Vianney Perchet -
2021 Poster: Online Matching in Sparse Random Graphs: Non-Asymptotic Performances of Greedy Algorithm »
Nathan Noiry · Vianney Perchet · Flore Sentenac -
2021 Poster: Decentralized Learning in Online Queuing Systems »
Flore Sentenac · Etienne Boursier · Vianney Perchet -
2020 Poster: Robustness of Community Detection to Random Geometric Perturbations »
Sandrine Peche · Vianney Perchet -
2020 Poster: Statistical Efficiency of Thompson Sampling for Combinatorial Semi-Bandits »
Pierre Perrault · Etienne Boursier · Michal Valko · Vianney Perchet -
2019 Poster: Categorized Bandits »
Matthieu Jedor · Vianney Perchet · Jonathan Louedec -
2017 Poster: Fast Rates for Bandit Optimization with Upper-Confidence Frank-Wolfe »
Quentin Berthet · Vianney Perchet -
2017 Spotlight: Fast Rates for Bandit Optimization with Upper-Confidence Frank-Wolfe »
Quentin Berthet · Vianney Perchet -
2016 Poster: Combinatorial semi-bandit with known covariance »
Rémy Degenne · Vianney Perchet