Table 1 Evaluation metrics of bi-stream CNN and conventional machine learning models
From: Bi-stream CNN Down Syndrome screening model based on genotyping array
Models | Evaluation metrics of different models | |||||
|---|---|---|---|---|---|---|
| Â | Accuracy | Precision | Recall | F-score | False-positive rate | False-negative rate |
Decision tree | 96.9(+-1.0)% | 94.1% | 95.4% | 94.6% | 2.2% | 8.0% |
Random forest | 97.1(+-0.7)% | 94.4% | 94.9% | 94.7% | 1.9% | 8.1% |
SVM | 96.7(+-0.9)% | 92.7% | 95.9% | 94.2% | 2.9% | 5.3% |
Bi-Stream CNN | 99.3(+-0.4)% | 99.2% | 98.4% | 99.3% | 0.6% | 1.1% |