Dear edward users, I’m trying to extend the model for “unsupervised learning by clustering data points” in the Mixture models tutorial ( http://edwardlib.org/tutorials/unsupervised ) My intention is to put the components as a parameter inside the model and to do inference in the number of components …

Dear @dustin and Edward users. Finally the code below works, i.e., DPM for a mixture of two dimensional Gaussians using the stick breaking method. I had some issues with the stick breaking function, but I’ve check out the code using TensorFlow: K = 5 # number of components to test alpha = Gamma(co…

DPM model for clustering

AngelBerihuete August 25, 2017, 7:27am 9

Hi @emilemathieu !

Have you seen the post written by @dustin about your issue?

Do you have any progress with DPM and Edward?

Cheers!

1 Like

Topic		Replies	Views
Dirichlet process mixture model for multivariate Bernoulli mixture	0	741	April 22, 2018
Gaussian mixture model (Dirichlet process)	0	889	June 4, 2018
Bernoulli Mixture Models on MNIST	3	1930	July 11, 2017
Hierarchical Dirichlet Model	1	1007	July 11, 2018
Implementing a factor potential in MCMC inference	1	878	November 9, 2017

DPM model for clustering

Related Topics