Recent searches
Search options
do the #statistics or #ML folks have any suggestions about packages (ideally in python) to train #Bayesian Gaussian mixture models? I was using the sklearn package but suspect there is something better
@BayesForDays not savvy on these models, but my knee-jerk reaction is that `sklearn` is usually the winner in this kind of contest unless you're doing something very domain-specific in a library that includes its own tools. Is there something in particular about `sklearn` that you're finding inadequate?
@BayesForDays@lingo.lol@SnoopJ I had some issues getting it to potentially assign *nothing* to a cluster, and inferring # of clusters on the basis of the data itself, which is maybe user error. But in any case I need something that knows to override my best guess