4 ms·
Note that there is some dubious stuff in the paper... Figure 3 B for example is deceptive... Drawing ellipses onto principal component analysis charts is ne
by londons_explore 5y ago
Note that there is some dubious stuff in the paper... Figure 3 B for example is deceptive... Drawing ellipses onto principal component analysis charts is never a rigorous way to know if something fits in a set... You are only visualising the first two components - and for all you know, the point you're querying might be miles away in the third axis that you aren't looking at. High dimensional data is high dimensional for a reason - conclusions you make by throwing away most of the data aren't robust!
Despite that, I still believe the overall conclusion of this paper is backed up by very strong evidence.