Safe Linear Stochastic Bandits

Authors: Kia Khezeli, Eilyan Bitar10202-10209

AAAI 2020 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental In this section, we conduct a simple numerical study to illustrate the qualitative features of the SEGE algorithm and compare it with the CLUCB algorithm introduced by (Kazerouni et al. 2017).
Researcher Affiliation Academia Kia Khezeli, Eilyan Bitar School of Electrical and Computer Engineering, Cornell University, Ithaca, NY, USA {kk839, eyb5}@cornell.edu
Pseudocode Yes Algorithm 1 SEGE Algorithm
Open Source Code No The paper does not provide any concrete access to source code (e.g., specific repository link, explicit code release statement, or code in supplementary materials) for the methodology described.
Open Datasets No The paper describes a simulated environment where data is generated for the experiments, rather than using a pre-existing publicly available dataset. Therefore, it does not provide concrete access information for a public dataset.
Dataset Splits No The paper describes a simulation setup rather than using an external dataset with predefined splits. It does not provide specific dataset split information (exact percentages, sample counts, citations to predefined splits, or detailed splitting methodology) needed to reproduce the data partitioning in the context of typical train/validation/test splits.
Hardware Specification No The paper does not provide specific hardware details (exact GPU/CPU models, processor types with speeds, memory amounts, or detailed computer specifications) used for running its experiments.
Software Dependencies No The paper does not provide specific ancillary software details (e.g., library or solver names with version numbers like Python 3.8, CPLEX 12.4) needed to replicate the experiment.
Experiment Setup Yes SEGE Algorithm. We set the parameters of the SEGE algorithm to c = 0.5, λ = 0.1, and ρ = ρ = 0.224. We generate the random exploration process according to Ut = x+ζt, where {ζt} t=1 is a sequence of IID random variables that are uniformly distributed on the unit circle. To enable a direct comparison between the SEGE and CLUCB algorithms, we restrict our attention to a summable sequence of risk levels that satisfy the conditions of Corollary 1. Specifically, we set the sequence of risk levels to δt = 6δ/(π2t2) for all stages t 1, where δ = 0.1.