Table 2.

Performance of ML algorithms for predicting ADAOO in the E280A pedigree. RMSE = root mean squared error, lower is better; MAE = mean absolute error, lower is better; R² = coefficient of determination, higher is better. Best results are shown in bold.

ML Algorithm	Performance Measure
	RMSE		R 2		MAE
	Training	Testing	Training	Testing	Training	Testing
glmboost	3.51	3.73	0.62	0.65	2.41	2.86
bstTree	3.67	6.75	0.59	0.08	3.00	4.52
gbm	4.90	6.68	0.27	0.09	3.86	4.52
glmnet	3.59	3.85	0.62	0.64	2.51	2.89
knn	4.53	6.35	0.39	0.05	3.56	4.13
mlp	6.30	6.62	0.07	0.43	5.64	5.78
qrf	1.35	7.24	0.95	0.03	0.69	4.65
rf	2.14	6.17	0.91	0.12	1.70	3.93
rpart	4.73	6.36	0.31	0.07	3.95	4.51
rpart1SE	4.18	5.89	0.46	0.18	3.35	4.11
rpart2	4.28	6.02	0.43	0.15	3.43	4.11
svmLinear	4.74	6.80	0.43	0.07	2.97	4.21
svmLinear2	4.74	6.80	0.43	0.07	2.97	4.21
svmPoly	3.46	7.30	0.66	0.14	1.86	5.13
svmRadial	5.21	6.50	0.35	0.02	3.43	3.96
treebag	4.26	6.02	0.45	0.16	3.47	4.20
xgbLinear	0.85	7.14	0.98	0.06	0.37	4.28
xgbTree	1.79	7.12	0.90	0.08	1.28	4.65