Notice: The reproducibility variables underlying each score are classified using an automated LLM-based pipeline, validated against a manually labeled dataset. LLM-based classification introduces uncertainty and potential bias; scores should be interpreted as estimates. Full accuracy metrics and methodology are described in [1].

Speeding Up Latent Variable Gaussian Graphical Model Estimation via Nonconvex Optimization

Authors: Pan Xu, Jian Ma, Quanquan Gu

NeurIPS 2017 | Venue PDF | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Experiments on both synthetic and genomic data demonstrate the superiority of our algorithm over the state-of-the-art algorithms and corroborate our theory. ... In this section, we present numerical results on both synthetic and real datasets to verify the theoretical properties of our algorithm, and compare it with the state-of-the-art methods.
Researcher Affiliation Academia Pan Xu Department of Computer Science University of Virginia Charlottesville, VA 22904 EMAIL Jian Ma School of Computer Science Carnegie Mellon University Pittsburgh, PA 15213 EMAIL Quanquan Gu Department of Computer Science University of Virginia Charlottesville, VA 22904 EMAIL
Pseudocode Yes Algorithm 1 Alternating Thresholded Gradient Descent (Alt GD) for LVGGM
Open Source Code No The paper states: 'The implementation of these two methods were downloaded from the authors website.' referring to competitor methods (PPA and ADMM), but does not provide any statement or link for the open-sourcing of their own method (Alt GD).
Open Datasets Yes We downloaded the gene expression data from cBio Portal2. Here we focused on 299 breast cancer related transcription factors (TFs) and estimated the regulatory relationships for each pair of TFs over two breast cancer subtypes: luminal and basal.
Dataset Splits Yes The regularization parameters for 1,1-norm and nuclear norm in PPA and ADMM and the tuning parameter r in our algorithm were selected by 4-fold cross-validation.
Hardware Specification Yes All numerical experiments were run in MATLAB R2015b on a laptop with Intel Core i5 2.7 GHz CPU and 8GB of RAM.
Software Dependencies Yes All numerical experiments were run in MATLAB R2015b on a laptop with Intel Core i5 2.7 GHz CPU and 8GB of RAM.
Experiment Setup Yes The step sizes of our method were set as = c1/(ฯƒmax 2) and 0 = c1ฯƒmin/(ฯƒmax 4) according to Theorem 4.5, where c1 = 0.25. The thresholding parameter s is set as c2s , where c2 > 1 was selected by 4-fold cross-validation.