Figure 4 : Neural-Network Approximate Bayesian Computation
posterior parameters estimated densities under the winning scenario
AfrDE-EurDE, for (A) the ACB and (B) the ASW populations. Median
posterior point estimates are indicated by the red vertical line, 95%
credibility intervals are indicated by the colored area under the
posterior curve (Table 2 ). All posterior parameter estimations
were conducted using 100,000 simulations under scenario AfrDE-EurDE, a
1% tolerance rate (1,000 simulations), 24 summary statistics, logit
transformation of all parameters, and four neurons in the hidden layer
(see Materials and Methods ). For all parameters separately,
densities are plotted with 1,000 points, a Gaussian kernel, and are
constrained to the prior limits. Posterior parameter densities are
indicated by a solid line; prior parameter densities are indicated by
black dotted lines.