Fig. 13: Neural Networks (BP) for the number of states (y1), duration (y2), number of alliances (y3), number of wars (y4).
From: Human civilization dynamics: why we have different civilization patterns in history

A1–A4 refer to the BP neural network outcomes of y1, y2, y3 & y4. B1–B4 are the probability density distribution of MSE in 100 iterations. C1–C4 plot the paired scatters.