Deep Learning for Population Genetic Inference
We evaluate the results in three ways, using the relative error of the estimates: |Nest − Ntrue|/Ntrue. First, for each test demography, we average the statistics for each dataset, and then run these values through the training network. Second, for each test demography, we run the datasets through the network one by one, then average the predictions. Finally, we perform the second procedure, but with only the regions we classified as neutral. We note that the most ancient size (N3) is always the least accurately estimated.