Accuracy and robustness of learning and inference in Bayesian networks
Jan R. J. Nunnink · UvA-DARE (University of Amsterdam) · 2008
factor φ k (X) factor φ k (Y ) p(Y |X) is adequate?adequate adequate ⇒ yes adequate inadequate ⇒ no inadequate adequate ⇒ no inadequate inadequate ⇒ yes Table 5.3: Localization rules if X and Y are binary variables.factor φ k (X) factor φ k (Y ) p(Y |X) is adequate?adequate adequate ⇒ yes adequate inadequate ⇒ unknown inadequate adequate ⇒ no inadequate inadequate ⇒ unknown