Table 3 Percentage error rates for the Bayesian Model (BM), GBM and Random Forest (RF) for the datasets defined in Table 2; for samples containing mixtures, samples containing single serotypes and all samples combined
From: A comparison of machine learning and Bayesian modelling for molecular serotyping
Dataset | BM | GBM | RF | ||||
|---|---|---|---|---|---|---|---|
D.36 | Mixtures | 18 | (15.2 - 21.0) | 5.6 | (4.0 - 7.4) | 10.0 | (7.9 - 12.3) |
| Â | Singles | 8.1 | (7.2 - 9.0) | 8.3 | (7.4 - 9.2) | 9.8 | (8.8 - 10.8) |
| Â | Combined | 9.7 | (8.8 - 10.6) | 7.9 | (7.1 - 8.7) | 9.8 | (8.0 - 9.7) |
D.73 | Mixtures | 21 | (18.4 - 23.7) | 19 | (16.5 - 21.6) | 21 | (18.4 - 23.7) |
| Â | Singles | 8.2 | (7.3 - 9.1) | 22.0 | (20.7 - 23.3) | 20.0 | (18.7 - 21.3) |
| Â | Combined | 10.8 | (9.9 - 11.7) | 21.4 | (20.2 - 22.6) | 20.2 | (19.1 - 21.4) |