Table 1 HS models performance evaluation in external validation datasets
From: Deep learning health space model for ordered responses
KNHANES(n = 32,140) | Ewha-Boramae (n = 862) | KARE (n = 3,199) | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
Average Health space index | Average Silhouette score | Calinski-Harabasz index | Davies Bouldin index | Average Health space index | Average Silhouette score | Calinski-Harabasz index | Davies Bouldin index | Average Health space index | Average Silhouette score | Calinski-Harabasz index | Davies Bouldin index | |
Deep Ordinal Neural Network | 0.74 | 0.09 | 49722.82 | 1.66 | 0.53 | -0.08 | 487.60 | 2.41 | 0.41 | 0.00 | 1475.77 | 3.34 |
Binary DNN (0 vs. 3) | 0.46 | -0.04 | 4988.67 | 4.37 | 0.43 | -0.21 | 121.16 | 4.16 | 0.40 | -0.15 | 399.87 | 4.42 |
Binary DNN (0 vs. 1 + 2 + 3) | 0.45 | -0.08 | 3386.12 | 5.83 | 0.42 | -0.28 | 74.70 | 9.57 | 0.36 | -0.26 | 267.53 | 4.3 |
Binary DNN (0 + 1 vs. 2 + 3) | 0.57 | -0.10 | 9693.17 | 10.59 | 0.51 | -0.20 | 157.85 | 5.45 | 0.38 | -0.10 | 759.1 | 2.68 |
Binary DNN (0 + 1 + 2 vs. 3) | 0.58 | -0.18 | 2681.57 | 8.17 | 0.49 | -0.29 | 62.25 | 13.56 | 0.29 | -0.11 | 141.49 | 13.59 |
Proportional Odds Model | 0.43 | -0.04 | 19134.58 | 5.35 | 0.47 | -0.10 | 319.78 | 9.67 | 0.40 | -0.02 | 882.99 | 3.19 |