Table 2 Accuracy of the proposed diet recommendation method in the original PROTEIN NAP database (without ChatGPT utilization).
From: AI nutrition recommendation using a deep generative model and ChatGPT
| Â | Macronutrients accuracy | Caloric difference | ||||
|---|---|---|---|---|---|---|
|  | Average | Fat | SFA | Protein | Carbohydrates | Mean ± Std (%) |
Healthy adults | 89.66 | 88.74 | 90.59 | – | – | 0.0 ± 0.0 |
Healthy adolescents | 88.97 | 87.52 | 90.42 | – | – | 0.0 ± 0.0 |
Healthy older adults | 88.22 | 87.02 | 89.42 | – | – | 0.0 ± 0.0 |
Adults who are overweight | 82.38 | 86.49 | 90.53 | 87.89 | 64.60 | 0.0 ± 0.0 |
Athletes | 86.37 | 84.88 | 86.41 | 84.05 | 90.15 | 0.0 ± 0.0 |
Adults who are obese | 89.60 | 85.08 | 91.68 | 88.48 | 93.16 | 0.0 ± 0.0 |
Adults with CVD | 81.15 | 60.68 | 91.48 | 86.75 | 85.71 | 0.0 ± 0.0 |
Adults with T2D | 86.58 | 84.24 | 86.59 | 86.73 | 82.30 | 0.0 ± 0.0 |
Adults with iron deficiency | 87.48 | 87.14 | 87.82 | – | – | 0.0 ± 0.0 |
Adults with a diet low in fruits/vegetables | 89.63 | 88.63 | 90.63 | – | – | 0.0 ± 0.0 |
Average | 87.00 | 84.04 | 89.55 | 86.78 | 83.18 | 0.0 ± 0.0 |
