Open ctb opened 7 years ago
@ekg suggested looking into variational autoencoding:
Or, if you're interested in finding your way back into the big tree, you could use VAE or similar dim. > reduction and work from that repr
(we already have t-SNE working in a notebook somewhere)
see cluster and cocluster too #700
https://stats.stackexchange.com/questions/3685/where-to-cut-a-dendrogram
http://www.sigmath.es.osaka-u.ac.jp/shimo-lab/prog/pvclust/
https://github.com/biocore/genome-subsampler/blob/master/genomesubsampler/prototypeSelection.py
things to provide code for --