Problem Dependent View on Structured Thresholding Bandit Problems

Authors: James Cheshire, Pierre Menard, Alexandra Carpentier

ICML 2021 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental In Appendix ?? we conduct some preliminary experiments to explore how our theoretical results translate in practice. All proofs are found in the Appendix.
Researcher Affiliation Academia 1Otto von Guericke University Magdeburg. Correspondence to: James Cheschire <james.cheschire@ovgu.de>.
Pseudocode Yes Algorithm 1 PD-MTB; Algorithm 2 Grad-Explore; Algorithm 3 PD-CTB
Open Source Code No The paper does not provide any explicit statements about the release of source code or links to a code repository.
Open Datasets No The paper focuses on a theoretical bandit problem setting and describes general problem formulations (e.g., 'K-armed bandit problem', 'unknown distribution νk'). It mentions 'preliminary experiments' but provides no specific details, names, links, or citations for any publicly available datasets used.
Dataset Splits No The paper does not specify any training, validation, or test dataset splits. The research is primarily theoretical with preliminary experiments mentioned but not detailed in terms of data splits.
Hardware Specification No The paper mentions 'preliminary experiments' but does not provide any specific details about the hardware used to conduct these experiments (e.g., GPU/CPU models, memory).
Software Dependencies No The paper mentions 'preliminary experiments' but does not provide specific version numbers for any software dependencies (e.g., programming languages, libraries, frameworks) used in these experiments.
Experiment Setup No The paper describes the theoretical setup of the Thresholding Bandit Problem and the parameters of its proposed algorithms (e.g., T1 and T2 derived from T and K). However, it does not provide specific experimental setup details such as hyperparameters, optimization settings, or system-level training configurations that would be used in an empirical implementation of the algorithms.