(a)
(b)
Figure 15: Prediction precision versus max generation of QPSO. (a) Ratio of training set = 60%; (b) ratio of training set = 80%.