Abstract 604 Table 2

AUC (95% CI) of all algorithms based on internal and external assessments

DevelopmentExternal Validation data sets
PROMISSEBronx, NYNYC, NYPisa, Italy
LASSO 0.77 (0.71,0.83)* 0.60 (0.46,0.73) 0.63 (0.47,0.80) 0.66 (0.53,0.79)
Support vector machine 0.77 (0.70,0.84)* 0.61 (0.47, 0.74) 0.58 (0.41, 0.74) 0.73 (0.63,0.83)
Random Forest 0.77 (0.71,0.83)* 0.68 (0.55,0.81) 0.57 (0.46,0.80) 0.67 (0.56, 0.79)
Super Learner 0.78 (0.72,0.84)* 0.66 (0.53,0.79) 0.63 (0.43,0.76) 0.71 (0.56, 0.81)

  • *Based on 5x10-fold cross-validation