AppliedStatisticsForNeuroscience icon indicating copy to clipboard operation
AppliedStatisticsForNeuroscience copied to clipboard
verified publisher icon charlesfrye

Clustering Tutorial Issues

Open charlesfrye opened this issue 8 years ago • 1 comments

The K-Means implementation is a hack -- it's just EM with small, isotropic covariance matrices. Need to do an honest-to-goodness K-Means implementation.

charlesfrye avatar May 04 '17 18:05 charlesfrye

While this would be nice, it works well enough with the "limit of EM" approach. I'll leave this open as a potential future fix.

charlesfrye avatar Oct 24 '17 18:10 charlesfrye