Methods for graph selection

for Gaussian graphical models:

1 -regularized neighborhood regression for Gaussian MRFs

(e.g., Meinshausen & Buhlmann, 2005; Wainwright, 2006, Zhao & Yu, 2006)

1 -regularized log-determinant

(e.g., Yuan & Lin, 2006; dAsprmont et al., e 2007; Friedman, 2008; Rothman et al., 2008; Ravikumar et al., 2008)

methods for discrete MRFs

exact solution for trees local testing various other methods

(Chow & Liu, 1967) (e.g., Spirtes et al, 2000; Kalisch & Buhlmann, 2008)

distribution ts by KL-divergence (Abeel et al., 2005) 1 -regularized log. regression (Ravikumar, W. & Laerty et al., 2008, 2010) approximate max. entropy approach and thinned graphical models (Johnson et al., 2007) neighborhood-based thresholding method (Bresler, Mossel & Sly, 2008)

information-theoretic analysis

pseudolikelihood and BIC criterion information-theoretic limitations

(Csiszar & Talata, 2006) (Santhanam & W., 2008, 2012)