Table 1 Prediction accuracy of the three top-performing models after rounds of optimization.
From: Blood Biochemistry Analysis to Detect Smoking Status and Quantify Accelerated Aging in Smokers
|  | No. of features | MAE (years) | r | ε-accuracy (ε = 10 years) | R 2 |
|---|---|---|---|---|---|
Age predictor trained on 23 features | 23 | 5.722 | 0.76 | 0.803 | 0.56 |
Age predictor trained on 20 features | 20 | 5.777 | 0.75 | 0.801 | 0.5376 |
Age predictor trained on 18 features | 18 | 5.898 | 0.75 | 0.802 | 0.55 |
Age predictor trained on 24 features | 24 | 5.61 | 0.78 | 0.82 | 0.578 |
Age predictor trained on 21 | 21 | 5.401 | 0.77 | 0.815 | 0.58 |
Age predictor trained on 19 features | 19 | 5.416 | 0.77 | 0.817 | 0.60 |
| Â | No. of features | Accuracy | Precision | Recall | F1 |
Smoking status classifier trained on 23 features | 23 | 0.829 | 0.754 | 0.606 | 0.673 |
Smoking status classifier trained on 20 features | 20 | 0.822 | 0.726 | 0.61 | 0.664 |
Smoking status classifier trained on 18 features | 18 | 0.82 | 0.708 | 0.603 | 0.638 |
