On generating random Gaussian graphical models
Research output: Contribution to journal › Journal article › peer-review
Structure learning methods for covariance and concentration graphs are often validated on synthetic models, usually obtained by randomly generating: (i) an undirected graph, and (ii) a compatible symmetric positive definite (SPD) matrix. In order to ensure positive definiteness in (ii), a dominant diagonal is usually imposed. In this work we investigate different methods to generate random symmetric positive definite matrices with undirected graphical constraints. We show that if the graph is chordal it is possible to sample uniformly from the set of correlation matrices compatible with the graph, while for general undirected graphs we rely on a partial orthogonalization method.
Original language | English |
---|---|
Journal | International Journal of Approximate Reasoning |
Volume | 125 |
Pages (from-to) | 240-250 |
Number of pages | 11 |
ISSN | 0888-613X |
DOIs | |
Publication status | Published - 2020 |
- Algorithm validation, Concentration graph, Covariance graph, Positive definite matrix simulation, Undirected graphical model
Research areas
Links
- https://arxiv.org/pdf/1909.01062.pdf
Accepted author manuscript
ID: 248192670