You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
There is a publication by Daniel Hsu and Sham Kakade (http://arxiv.org/abs/1206.5766) that describes how to estimate certain GMMs using the Method of Moments and tensor decomposition. This as appears to be interesting, as a MLE tends to end up in local optima.
I implemented this approach for my master thesis and if there is interest in such a contribution, I could port my code for usage with GaussianMixtures.jl.
Just let me know!
The text was updated successfully, but these errors were encountered:
There's also a more recent paper that introduces computationally efficient algorithms for method of moments estimation for Gaussian mixtures: https://arxiv.org/abs/2202.06930.
There is a publication by Daniel Hsu and Sham Kakade (http://arxiv.org/abs/1206.5766) that describes how to estimate certain GMMs using the Method of Moments and tensor decomposition. This as appears to be interesting, as a MLE tends to end up in local optima.
I implemented this approach for my master thesis and if there is interest in such a contribution, I could port my code for usage with GaussianMixtures.jl.
Just let me know!
The text was updated successfully, but these errors were encountered: