Table 2 Model performances measured by average AUC score (mean ± std) for multi-task predictions on real datasets.

From: Recurrent Neural Networks for Multivariate Time Series with Missing Values

Models ICD-9 20 Tasks on MIMIC-III Dataset All 4 Tasks on PhysioNet Dataset
GRU-Mean 0.7070 ± 0.001 0.8099 ± 0.011
GRU-Forward 0.7077 ± 0.001 0.8091 ± 0.008
GRU-Simple 0.7105 ± 0.001 0.8249 ± 0.010
GRU-CubicSpline 0.6372 ± 0.005 0.7451 ± 0.011
GRU-MICE 0.6717 ± 0.005 0.7955 ± 0.003
GRU-MF 0.6805 ± 0.004 0.7727 ± 0.003
GRU-PCA 0.7040 ± 0.002 0.8042 ± 0.006
GRU-MissForest 0.7115 ± 0.003 0.8076 ± 0.009
Proposed GRU-D 0.7123 ± 0.003 0.8370 ± 0.012