Skip to main content
. Author manuscript; available in PMC: 2020 Nov 18.
Published in final edited form as: Proc Int Conf Mach Learn Appl. 2018 Jan 18;2017:304–308. doi: 10.1109/icmla.2017.0-143

Table 3.

Mean absolute error (MAE) between observed and predicted scores across test folds, using machine learning to optimally combine per-item deletion and substitution errors.

Model 1-best Stim-Rescore
Ridge 0.052 0.040
Lasso 0.06 0.041
Linear SVR 0.066 0.058