Node-Based Learning of Multiple Gaussian Graphical Models

Authors: Karthik Mohan, Palma London, Maryam Fazel, Daniela Witten, Su-In Lee

JMLR 2014 | Venue PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Our proposal is illustrated on synthetic data, a webpage data set, and a brain cancer gene expression data set. Sections 6 and 7 are titled 'Simulation Study' and 'Real Data Analysis' respectively, providing empirical validation.
Researcher Affiliation Academia All authors are affiliated with the University of Washington, across different departments. Email domains are @uw.edu and @cs.washington.edu, which are academic.
Pseudocode Yes The paper contains 'Algorithm 1: ADMM algorithm for the PNJGL optimization problem (6)' and 'Algorithm 2: ADMM algorithm for the CNJGL optimization problem (7)'.
Open Source Code Yes Matlab code implementing CNJGL and PNJGL is available at http://faculty.washington. edu/mfazel/, http://www.biostat.washington.edu/~dwitten/software.html, and http://suinlee.cs.washington.edu/software.
Open Datasets Yes Our proposal is illustrated on synthetic data, a webpage data set, and a brain cancer gene expression data set... We applied PNJGL and CNJGL to the university webpages data set from the World Wide Knowledge Base project at Carnegie Mellon University. This data set was pre-processed by Cardoso-Cachopo (2009)... We applied the proposed methods to a publicly available gene expression data set... We downloaded the raw data in .CEL format from the The Caner Genome Atlas (TCGA) website.
Dataset Splits Yes We performed 5-fold cross-validation of the log-likelihood... for PNJGL, FGL, CNJGL, GGL, and GL, using a range of tuning parameters.
Hardware Specification Yes In a small example with p = 30, run on an Intel Xeon X3430 2.4Ghz CPU, the interior point method (using cvx, which calls Sedumi) takes 7 minutes to run, while the ADMM algorithm for PNJGL, coded in Matlab, takes only 0.58 seconds.
Software Dependencies No The paper mentions 'modeling environment cvx', 'conic interior-point solvers such as Se Du Mi or SDPT3', 'Matlab', 'SFNG functions in Matlab (George, 2007)', and 'Com Bat (Johnson and Li, 2006)'. However, no specific version numbers are provided for these software components.
Experiment Setup Yes We set µ = 5, ρ = 0.5 and tmax = 1000 in the PNJGL and CNJGL algorithms. In our implementation of these algorithms, the stopping criterion for the inner loop (corresponding to a fixed ρ) is ... ϵ is a tolerance that is chosen in our experiments to equal 10 4. ... Initialize: Primal variables to the identity matrix and dual variables to the zero matrix. ... PNJGL was performed with λ1 = 0 and λ2 = 2. ... CNJGL was performed with λ1 = 13 and λ2 = 410.