Figure 2.
Boxplots of pointwise mean absolute error (PMAE) achieved by the models. (A) The accuracy of MWen over all test samples and (C) with respect to prediction time points. (B) The accuracy of MPark over all test samples and (D) with respect to prediction time points. ∗0.01 < P ≤ 0.05; ∗∗∗∗P ≤ 0.0001. MPark = recurrent neural network method from Park et al; MWen = convolutional neural network method from Wen et al; ns = not significant; ROP = rate of progression.