Testing the Feasibility of Linear Programs with Bandit Feedback

Authors: Aditya Gangrade, Aditya Gopalan, Venkatesh Saligrama, Clayton Scott

ICML 2024 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental We conclude the paper by describing a heuristic implementation of EOGT, and its behaviour on the simple case of testing the feasibility of two linear constraints over the unit ball. The simulation is run for d [2 : 10]. In the varying Γ scenario, we fix d = 4, and impose the constraints x1 1/2 Γ for the feasible setting, and the constraints x1 Γ, x1 Γ in the infeasible case. The range Γ [0.2, 1] is studied at a grid of scale 0.1. Figure 2. Behaviour of the stopping time as d is varied for fixed Γ = 1/2 (left) and Γ is varied for fixed d = 4 (right) over the unit ball with m = 2. Averages and one-sigma error bars over 50 runs are reported.
Researcher Affiliation Academia 1Department of Electrical Computer Engineering, Boston University 2Department of Electrical Engineering and Computer Science, University of Michigan 3Department of Electrical Communication Engineering, Indian Institute of Science. Correspondence to: Aditya Gangrade <gangrade@bu.edu>.
Pseudocode Yes Algorithm 1 Ellipsoidal Optimistic-Greedy Test (EOGT) Algorithm 2 Tempered EOGT (T-EOGT)
Open Source Code No No explicit statement about the release of source code or a link to a code repository for the methodology was found. The paper only states 'The code was implemented in MATLAB' in the simulations section.
Open Datasets No The paper defines specific instances for simulation (e.g., 'the feasible instance x1 0, x2 0' or 'x1 1/2 Γ'), which are custom-defined for the experiments, not publicly available datasets with explicit access information.
Dataset Splits No The paper describes running simulations on defined instances but does not mention specific training, validation, or test dataset splits or cross-validation.
Hardware Specification Yes The code was implemented in MATLAB, and executed on a consumer grade Ryzen 5 CPU, with no multithreading, and took about 4 hours to run.
Software Dependencies No The paper states 'The code was implemented in MATLAB' but does not provide a specific version number for MATLAB or any other software dependencies.
Experiment Setup Yes Throughout, the feedback noise is independent Gaussian with standard deviation σ = 0.1 (the value of σ is used in the confidence radii, and in general, τ should be proportional to σ2). The parameter δ is set to 0.1, N = 1, and all results are averaged over 50 runs.