Graphical Nonconvex Optimization via an Adaptive Convex Relaxation

Authors: Qiang Sun, Kean Ming Tan, Han Liu, Tong Zhang

ICML 2018 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental We show via numerical studies that the proposed estimator outperforms other popular methods for estimating Gaussian graphical models.
Researcher Affiliation Collaboration 1Department of Statistical Sciences, University of Toronto, Toronto, Ontario, Canada 2School of Statistics, University of Minnesota, Minneapolis, MN, USA 3Tencent AI Lab, Tencent Technology, Shenzhen, China.
Pseudocode Yes Algorithm 1 A sequential convex approximation for the graphical nonconvex optimization.
Open Source Code No The paper states that 'Algorithm 1 can be implemented using existing R packages such as glasso.' but does not provide any link or explicit statement about releasing its own source code.
Open Datasets No The paper states, 'Finally, we generate the data according to X(1), . . . , X(n) i.i.d. N(0, Σ),' indicating that synthetic data was used, not a publicly available dataset.
Dataset Splits No The paper describes generating synthetic data and evaluating methods, but it does not specify any training, validation, or test dataset splits or cross-validation setups. It states, 'We present the results averaged over 100 data sets for each of the two simulation settings'.
Hardware Specification No The paper does not specify any hardware details such as GPU/CPU models, memory, or specific computing infrastructure used for the experiments.
Software Dependencies No The paper mentions that 'Algorithm 1 can be implemented using existing R packages such as glasso,' but it does not provide specific version numbers for these software dependencies.
Experiment Setup Yes For our proposal, we consider T = 4 iterations with the SCAD penalty proposed by Fan & Li (2001)... In all of our simulation studies, we pick γ = 2.1.