Distributed Multi-Task Learning for Stochastic Bandits with Context Distribution and Stage-wise Constraints
Not on the map. Why?
Abstract (arXiv)
We present conservative distributed multi-task learning in stochastic linear contextual bandits with heterogeneous agents. This extends conservative linear bandits to a distributed setting where M agents tackle different but related tasks while adhering to stage-wise performance constraints. The exact context is unknown, and only a context distribution is available to the agents as in many practical applications that involve a prediction mechanism to infer context, such as stock market prediction and weather forecast. We propose a distributed upper confidence bound (UCB) algorithm, DiSC-UCB. Our algorithm constructs a pruned action set during each round to ensure the constraints are met. Additionally, it includes synchronized sharing of estimates among agents via a central server using well-structured synchronization steps. We prove the regret and communication bounds on the algorithm. We extend the problem to a setting where the agents are unaware of the baseline reward. For this setting, we provide a modified algorithm, DiSC-UCB2, and we show that the modified algorithm achieves the same regret and communication bounds. We empirically validated the performance of our algorithm on synthetic data and real-world Movielens-100K data.
Key facts. Cited by 2 papers in the map (data through 7 Oct 2026). Builds on 29 of its 53 references that are in the map. First posted on arXiv 21 Jan 2024.
Builds on
29 of 53 references.
References · 29
- 2019Linear Stochastic Bandits Under Safety Constraintscited by 90
- 2020High-Dimensional Sparse Linear Banditscited by 61
- 2021Federated Linear Contextual Banditscited by 61
- 2016Conservative Banditscited by 58
- 2016Conservative Contextual Linear Banditscited by 56
- 2015Bayesian Incentive-Compatible Bandit Explorationcited by 53
- 2020Impact of Representation Learning in Linear Banditscited by 51
- 2016Distributed Clustering of Linear Bandits in Peer to Peer Networkscited by 50
- 2017Multi-Task Learning for Contextual Banditscited by 47
- 2021Near-optimal Representation Learning for Linear Bandits and Linear RLcited by 45
- 2020Information Directed Sampling for Linear Partial Monitoringcited by 39
- 2020Stage-wise Conservative Linear Banditscited by 28
17 more
- 2019Safe Linear Stochastic Banditscited by 27
- 2022Multi-task Representation Learning with Stochastic Linear Banditscited by 20
- 2019Stochastic Bandits with Context Distributionscited by 18
- 2021Safe Online Convex Optimization with Unknown Linear Safety Constraintscited by 13
- 2022Provable General Function Class Representation Learning in Multitask Bandits and MDPscited by 12
- 2015Distributed Multitask Learningcited by 11
- 2023Linear Partial Monitoring for Sequential Decision-Making: Algorithms, Regret Bounds and Applicationscited by 11
- 2021Analysis of Thompson Sampling for Partially Observable Contextual Multi-Armed Banditscited by 10
- 2022Multi-Player Multi-Armed Bandits with Finite Shareable Resources Arms: Learning Algorithms & Applicationscited by 10
- 2017Contextual Linear Bandits under Noisy Features: Towards Bayesian Oraclescited by 8
- 2021Robust Bandit Learning with Imperfect Contextcited by 8
- 2022Stochastic Conservative Contextual Linear Banditscited by 6
- 2018Conservative Exploration using Interleavingcited by 4
- 2023Byzantine-Resilient Decentralized Multi-Armed Banditscited by 3
- 2021Decentralized Upper Confidence Bound Algorithms for Homogeneous Multi-Agent Multi-Armed Banditscited by 2
- 2022Distributed Stochastic Bandit Learning with Delayed Context Observationcited by 2
- 2023Federated Learning for Heterogeneous Bandits with Unobserved Contextscited by 1
Built on by
Citing papers per year: 2024: 1; 2025: 0; 2026: 1.
2 papers.