Figure 12.
1-D marginal posterior pdf (blue line), prior pdf (red) for the 22 VP parameters (Tables 1 and 2). The same trained networks as used for Fig. 10 are applied to ten different EHB input patterns (Fig. 9). The regions between discontinuities are displayed on different rows. For every region, depth decreases from left to right in the figure. Note that the range of the vertical axis, that is, normalized probability, differs between rows.