A Lyapunov-Based Methodology for Constrained Optimization with Bandit Feedback

Authors: Semih Cayci, Yilin Zheng, Atilla Eryilmaz3716-3723

AAAI 2022 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Simulations We implement both Ly Off and Ly On algorithms for K = 2 arms with Bernoulli distributed rewards, costs, and penalties. [...] Figure 1a, 1b, and 1c show the simulation results (averaged over 10^4 runs) with v0 = 1 and δ0 = 0.5 for Ly Off and Ly On algorithms.
Researcher Affiliation Academia Semih Cayci,1*Yilin Zheng,2*Atilla Eryilmaz2 1Coordinated Science Laboratory, University of Illinois at Urbana-Champaign, Urbana, IL 61801. 2 Department of Electrical and Computer Engineering, The Ohio State University, Columbus, OH 43210 scayci@illinois.edu, zheng.1443@osu.edu, eryilmaz.2@osu.edu
Pseudocode Yes Algorithm 1: Ly Off Algorithm and Algorithm 2: Ly On Algorithm
Open Source Code No The paper does not provide any statement or link regarding the availability of its source code.
Open Datasets No The paper describes a synthetic simulation environment with specified parameters for Bernoulli distributed rewards, costs, and penalties for 2 arms, rather than using or providing a publicly available dataset.
Dataset Splits No The paper describes simulation parameters and observed outcomes but does not refer to standard dataset splits like training, validation, or test sets.
Hardware Specification No The paper does not provide any specific details about the hardware used to run the simulations or experiments.
Software Dependencies No The paper does not provide specific software dependencies with version numbers.
Experiment Setup Yes Figure 1a, 1b, and 1c show the simulation results (averaged over 10^4 runs) with v0 = 1 and δ0 = 0.5 for Ly Off and Ly On algorithms. [...] Figure 1d, 1e, and 1f show that we can indeed select specific v0 and δ0 values such that the constraint-violation becomes negative when B is sufficiently large.