Skip to main content
. 2020 Sep 7;2020:3729715. doi: 10.34133/2020/3729715

Table 2.

Cross-validation (CV) results for training of beginning of stem elongation support vector machine. Results are root mean squared errors (RMSEs) in growing degree days (GDDs) and relative root mean squared errors (rRMSEs).

Training Test Test type RMSE rRMSE
All data All data 10-fold CV 52.6 10.9%
Genotype subset Unseen genotypes 10-fold CV 51.7 10.6%
Year-site subset Unseen year-sites 3-fold CV 85.3 17.7%
Genotype subset in FIP18/FIP19 subset Unseen genotypes in Delley19 10-fold CV 89.8 18.5%
Genotype subset in FIP18/Delley19 subset Unseen genotypes in FIP19 10-fold CV 80.8 18.4%
Genotype subset in FIP19/Delley19 subset Unseen genotypes in FIP18 10-fold CV 79.7 16.7%