Optimal Shrinkage of Singular Values Under Random Data Contamination

Authors: Danny Barash, Matan Gavish

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

Reproducibility Variable Result LLM Response
Research Type Experimental Simulations were performed to verify the correctness of our main results. Figure 2: Left: empirical validation of the predicted critical signal level (Simulation 1). Right: Empirical validation of the optimal shrinker shape (Simulation 3).
Researcher Affiliation Academia Danny Barash School of Computer Science and Engineering Hebrew University Jerusalem, Israel danny.barash@mail.huji.ac.il Matan Gavish School of Computer Science and Engineering Hebrew University Jerusalem, Israel gavish@cs.huji.ac.il
Pseudocode No The paper describes the proposed algorithms mathematically and textually, but it does not contain any structured pseudocode or clearly labeled algorithm blocks.
Open Source Code Yes The full Matlab code that generated the figures in this paper and in the Supporting Information is permanently available at https://purl.stanford.edu/kp113fq0838.
Open Datasets No The paper describes generating synthetic data matrices for simulations ("several independent data matrices were generated") based on a signal model, but it does not use a publicly available or open dataset for which access information is provided.
Dataset Splits No The paper describes a simulation-based approach and evaluation of theoretical predictions, but it does not specify explicit train/validation/test dataset splits or cross-validation methods for experimental reproduction.
Hardware Specification No The paper does not provide any specific hardware details (e.g., GPU/CPU models, memory) used for running its experiments or simulations.
Software Dependencies No The paper mentions 'Matlab code' but does not provide specific version numbers for Matlab or any other ancillary software dependencies required to replicate the experiments.
Experiment Setup Yes Figure 1: Left: Optimal shrinker for additive noise and missing-at-random contamination. Right: Phase plane for critical signal levels, see Section 6, Simulation 2. (β=0.3 β=0.6 β=1 shown in graph) and Figure 2, left panel, shows the amount of data singular values yi above xcritical(λ ), as a function of the fraction of missing values κ.