Table 8.
Validation results for the GBM model using the herd splitting method
| BCS | Herd 1 | Herd 2 | Herd 3 | |||
|---|---|---|---|---|---|---|
| Exact | 0.5-unit DEV | Exact | 0.5-unit DEV | Exact | 0.5-unit DEV | |
| 1.5 | 3 | 67 | 0 | 90 | 0 | 78 |
| 2.0 | 25 | 86 | 42 | 98 | 29 | 88 |
| 2.5 | 51 | 93 | 63 | 99 | 58 | 92 |
| 3.0 | 62 | 94 | 52 | 97 | 33 | 95 |
| 3.5 | 49 | 93 | 38 | 97 | 29 | 81 |
| 4.0 | 19 | 88 | 25 | 88 | 12 | 69 |
| 4.5 | 0 | 100 | 0 | 40 | 0 | 32 |
| WAvg | 42.2 | 89.8 | 48.0 | 95.9 | 35.9 | 86.6 |
| R 2 | 0.49 | 0.67 | 0.45 | |||
| RMSE | 0.32 | 0.33 | 0.35 | |||
Accuracy of the BCS categories is presented in percentage. GBM = Gradient Boosting Machine, BCS = Body Condition Score, R2 = R-square, RMSE = Root Mean Square Error, Exact = exact score, DEV = deviation, WAvg = weighted average by frequency.