Can we get the value of the latent variable for each training example while fitting a Gaussian mixture model by performing kmeans on the data set ? Further can we then estimate the other parameters of the GMM like we do in the case of Gaussian discriminant analysis ?
Add a comment
|