Skip to main content

Table 2 Accuracy measures for BNCs

From: Evaluation of Bayesian classifiers in asthma exacerbation prediction after medication discontinuation

Bayesian network classifier Accuracy Sensitivity Specificity
Naive Bayes 75.38% (89.2%) 72.72% (54.54%) 75.9% (96.2%)
TAN (BIC) 76.92% (86.15%) 72.72% (54.54%) 77.77% (92.6%)
TAN (AIC) 73.84% (84.61%) 81.81% (36.36%) 72.22% (94.44%)
TAN (log-likelihood) 75.38% (84.61%) 63.63% (27.27%) 77.77% (96.3%)
TAN (HC) 76.92% (86.15%) 72.72% (54.54%) 77.77% (92.6%)
FSSJ 53.84% (86.15%) 81.81% (36.36%) 48.15% (96.3%)
BSEJ 93.84% (89.2%) 90.9% (54.54%) 94.44% (96.3%)