Table 2:
MIMIC II results across frequent (S), few-shot (F), and zero-shot (Z) groups. We mark prior methods for MIMIC datasets that we implemented with a *.
| S | F | Z | Harmonic Average | |||||
|---|---|---|---|---|---|---|---|---|
| R@5 | R@10 | R@5 | R@10 | R@5 | R@10 | R@5 | R@10 | |
| Random | 0.000 | 0.000 | 0.000 | 0.000 | 0.011 | 0.032 | 0.000 | 0.000 |
| Logistic (Vani et al., 2017) * | 0.137 | 0.247 | 0.001 | 0.003 | - | - | - | - |
| CNN (Baumeletal.,2018)* | 0.138 | 0.250 | 0.050 | 0.082 | - | - | - | - |
| ACNN (Mullenbach et al., 2018) * | 0.138 | 0.255 | 0.046 | 0.081 | - | - | - | - |
| Match-CNN (Rios and Kavuluru, 2018) | 0.137 | 0.247 | 0.031 | 0.042 | - | - | - | - |
| ESZSL + W2V | 0.074 | 0.119 | 0.008 | 0.017 | 0.080 | 0.172 | 0.020 | 0.041 |
| ESZSL + W2V 2 | 0.050 | 0.086 | 0.025 | 0.044 | 0.103 | 0.189 | 0.043 | 0.076 |
| ESZSL + GRALS | 0.135 | 0.238 | 0.081 | 0.123 | 0.085 | 0.136 | 0.095 | 0.152 |
| ZACNN | 0.135 | 0.245 | 0.103 | 0.149 | 0.147 | 0.221 | 0.128 | 0.205 |
| ZAGCNN | 0.135 | 0.247 | 0.130 | 0.185 | 0.269 | 0.362 | 0.160 | 0.246 |