Fig. 2. (Color online) Performance of the machine learning model trained with the dataset generated according to the rule shown in Fig.
1. The figure labels (a–f) correspond to those presented in Fig.
1: (a–c) Samples independent of
V. (d–f) Samples dependent on
V. The conditions for
ε are (a) and (e) constant, (b) and (f) uniform, (c) random, and (d) sample independent. The colors blue and orange represent
R2 and
χ2, respectively. In (a), (b), and (d),
χ2 is fitted on a semi-log scale.
