A partial orthogonalization method for simulating covariance and concentration graph matrices
Publikation: Bidrag til bog/antologi/rapport › Konferencebidrag i proceedings › Forskning › fagfællebedømt
Standard
A partial orthogonalization method for simulating covariance and concentration graph matrices. / Córdoba, Irene ; Varando, Gherardo; Bielza, Concha ; Larranaga, Pedro .
Proceedings of the 9th International Conference on Probabilistic Graphical Models. red. / Václav Kratochvíl; Milan Studený. PMLR, 2018. s. 61-72 (Proceedings of Machine Learning Research, Bind 72).Publikation: Bidrag til bog/antologi/rapport › Konferencebidrag i proceedings › Forskning › fagfællebedømt
Harvard
APA
Vancouver
Author
Bibtex
}
RIS
TY - GEN
T1 - A partial orthogonalization method for simulating covariance and concentration graph matrices
AU - Córdoba, Irene
AU - Varando, Gherardo
AU - Bielza, Concha
AU - Larranaga, Pedro
PY - 2018
Y1 - 2018
N2 - 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. However, the link strengths in the resulting graphical model, determined by off-diagonal entries in the SPD matrix, are in many scenarios extremely weak. Recovering the structure of the undirected graph thus becomes a challenge, and algorithm validation is notably affected. In this paper, we propose an alternative method which overcomes such problem yet yields a compatible SPD matrix. We generate a partially row-wise-orthogonal matrix factor, where pairwise orthogonal rows correspond to missing edges in the undirected graph. In numerical experiments ranging from moderately dense to sparse scenarios, we obtain that, as the dimension increases, the link strength we simulate is stable with respect to the structure sparsity. Importantly, we show in a real validation setting how structure recovery is greatly improved for all learning algorithms when using our proposed method, thereby producing a more realistic comparison framework.
AB - 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. However, the link strengths in the resulting graphical model, determined by off-diagonal entries in the SPD matrix, are in many scenarios extremely weak. Recovering the structure of the undirected graph thus becomes a challenge, and algorithm validation is notably affected. In this paper, we propose an alternative method which overcomes such problem yet yields a compatible SPD matrix. We generate a partially row-wise-orthogonal matrix factor, where pairwise orthogonal rows correspond to missing edges in the undirected graph. In numerical experiments ranging from moderately dense to sparse scenarios, we obtain that, as the dimension increases, the link strength we simulate is stable with respect to the structure sparsity. Importantly, we show in a real validation setting how structure recovery is greatly improved for all learning algorithms when using our proposed method, thereby producing a more realistic comparison framework.
M3 - Article in proceedings
T3 - Proceedings of Machine Learning Research
SP - 61
EP - 72
BT - Proceedings of the 9th International Conference on Probabilistic Graphical Models
A2 - Kratochvíl, Václav
A2 - Studený, Milan
PB - PMLR
T2 - 9th International Conference on Probabilistic Graphical Models
Y2 - 11 September 2018 through 14 September 2018
ER -
ID: 215089091