Estimating Sparse Graphical Models: Insights Through Simulation

• Author / Creator

  Zhu, Yunan

• Graphical models are frequently used to explore networks among a set of variables. Several methods for estimating sparse graphs have been proposed and their theoretical properties have been explored. There are also several selection criteria to select the optimal estimated models. However, their practical performance has not been studied in detail. In this work, several estimation procedures (glasso, bootstrap glasso, adptive lasso, SCAD, DP-glasso and Huge) and several selection criteria (AIC, BIC, CV, ebic, ric and stars) are compared under various simulation settings, such as different dimensions or sample sizes, different types of data, and different sparsity levels of the true model structures. Then we use several evaluation criteria to compare the optimal estimated models and discuss in detail the superiority and deficiency of each combination of estimating methods and selection criteria.

• Subjects / Keywords

  • Graphical model
  • Glasso
  • Bootstrap
  • Penalized log-likelihood
  • Estimate evaluation

• Graduation date

  Fall 2015

• Type of Item

  Thesis

• Degree

  Master of Science

• DOI

  https://doi.org/10.7939/R3F766G1T

• License

  This thesis is made available by the University of Alberta Libraries with permission of the copyright owner solely for non-commercial purposes. This thesis, or any portion thereof, may not otherwise be copied or reproduced without the written consent of the copyright owner, except to the extent permitted by Canadian copyright law.