Figure 10
From: Constructing neural networks with pre-specified dynamics
Dependence of correlation structure on the optimised measure. (a) Matrix showing correlations between measures along the evolutionary process. Row labels indicate the measure employed as fitness function, and collect correlation coefficients (CC) between the optimised measure and the other measures. A symmetric matrix would imply that correlations are completely independent of which measure of the correlated pair of measures was the optimised one. (b) Entry (row, col) in matrix shown in (a), plotted in the x axis, against entry (col, row) in the y axis. Spearman correlation \(\rho =0.69\), \(p=8.10^{-5}\). A symmetric matrix in (a) would result in points lying in a line of slope = 1.
