Post #2713954

@robjohnnoble@ecoevo.social This is so cool! Have you considered applying this approach to index design to co-phylogenies? I've been working on supervised learning for co-phylogeny classification problems for a long time, and the biggest hurdle has been establishing some kind of ground-thruth for the classifier output. Metrics on trees are often frustratingly unrelated to the simulation parameters used to make the trees in the first place, which bodes poorly for their application to trees inferred from data.