Distributed Bandit Learning: Near-Optimal Regret with Efficient Communication
Authors: Yuanhao Wang, Jiachen Hu, Xiaoyu Chen, Liwei Wang
ICLR 2020 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Theoretical | The paper focuses on designing algorithms (protocols), proving theoretical guarantees (regret bounds, communication costs), and deriving lower bounds. There are no mentions of datasets, experimental setups, performance metrics from actual runs, or comparisons on real-world or simulated data. It's purely theoretical. For example, in Section 1.2 'OUR CONTRIBUTION', it states: 'In both settings, we present communication-efficient protocols that achieve near-optimal regret. Our results are summarized in Table 1.' and 'We also prove the following lower bound:'. Further, sections like '3.3 REGRET AND COMMUNICATION EFFICIENCY OF DEMAB' and the entire Appendix (Sections C, D, E, G, H) are dedicated to mathematical proofs and theoretical analysis. |
| Researcher Affiliation | Academia | Yuanhao Wang Institute for Interdisciplinary Information Sciences, Tsinghua University yuanhao-16@mails.tsinghua.edu.cn; Jiachen Hu* School of Electronics Engineering and Computer Science, Peking University Nick H@pku.edu.cn; Xiaoyu Chen Key Laboratory of Machine Perception, MOE, School of EECS, Peking University cxy30@pku.edu.cn; Liwei Wang Key Laboratory of Machine Perception, MOE, School of EECS Center for Data Science, Peking University wanglw@cis.pku.cn |
| Pseudocode | Yes | The paper includes multiple clearly labeled pseudocode blocks, for example: 'Protocol 1: Distributed Elimination for Multi-Armed Bandits (DEMAB)' in Section 3.2, 'Protocol 2: Distributed Elimination for Linear Bandits (DELB)' in Section 4.2, and 'Protocol 3: Distributed Linear UCB (Dis Lin UCB)' in Section 4.4. Additional protocols are detailed in Appendix B. |
| Open Source Code | No | The paper does not contain any statements or links indicating that source code for the described methodologies is publicly available or will be released. |
| Open Datasets | No | The paper is theoretical, focusing on algorithm design and mathematical analysis of regret and communication costs. It does not describe any empirical experiments or the use of datasets for training, validation, or testing. |
| Dataset Splits | No | The paper is theoretical and does not include any empirical experiments. Consequently, there are no mentions of dataset splits (e.g., training, validation, test percentages or counts, or citations to predefined splits) that would be needed for reproducibility. |
| Hardware Specification | No | The paper is theoretical and focuses on algorithm design and mathematical analysis. It does not describe any empirical experiments or their computational execution, and therefore, no specific hardware specifications are mentioned. |
| Software Dependencies | No | The paper is theoretical and focuses on algorithm design and mathematical analysis. It does not describe any specific software implementations or dependencies with version numbers. |
| Experiment Setup | No | The paper is purely theoretical, presenting algorithms and their performance bounds mathematically. It does not describe any empirical experimental setups, hyperparameters, or training configurations. |