Approximate Supermodularity Bounds for Experimental Design

Authors: Luiz Chamon, Alejandro Ribeiro

NeurIPS 2017 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental In this section, we illustrate the previous results in some numerical examples. To do so, we draw the elements of Ae from an i.i.d. zero-mean Gaussian random variable with variance 1/p and p = 20. The noise {ve} are also Gaussian random variables with Re = σ2 v I. We take σ2 v = 10 1 in high SNR and σ2 v = 10 in low SNR simulations.
Researcher Affiliation Academia Luiz F. O. Chamon and Alejandro Ribeiro Electrical and Systems Engineering University of Pennsylvania {luizf,aribeiro}@seas.upenn.edu
Pseudocode No The paper describes a greedy solution iteratively in text and with an equation, but does not present it as a formally structured pseudocode or algorithm block.
Open Source Code No The paper does not provide an explicit statement or a link indicating that the source code for the methodology described in this paper is openly available.
Open Datasets Yes In the following example, we use a subset of the Each Movie dataset [23] to illustrate how greedy experimental design can be applied to address this problem. ... [23] Digital Equipment Corporation, Each Movie dataset, http://www.gatsby.ucl.ac.uk/~chuwei/data/Each Movie/.
Dataset Splits No The paper mentions 'a training and test set containing 9000 and 3000 users respectively' but does not explicitly state a validation split or its size.
Hardware Specification No The paper does not provide specific hardware details (e.g., CPU/GPU models, memory) used for running the experiments.
Software Dependencies No The paper does not list specific software dependencies with version numbers needed to replicate the experiment.
Experiment Setup Yes To do so, we draw the elements of Ae from an i.i.d. zero-mean Gaussian random variable with variance 1/p and p = 20. The noise {ve} are also Gaussian random variables with Re = σ2 v I. We take σ2 v = 10 1 in high SNR and σ2 v = 10 in low SNR simulations. The experiment pool contains #E = 200 experiments. ... Starting with A-optimal design, we display the bound from Theorem 3 in Figure 1a for multivariate measurements of size ne = 5 and designs of size k = 40. ... We also let H = I and take a non-informative prior θ = 0 and Rθ = σ2 θI with σ2 θ = 100.