I have uploaded a plot to the data-science channel at Slack showing my attempt to reproduce the results in Negoita et al, 1810.04009 I hope somebody else is working on this. It seems my networks are not as good as Negoita et al and it would be good if somebody could show me how to do better. Richard