# [Topic-models] LDA paper by Blei - some more questions

Mateusz Berezecki mateuszb at gmail.com
Fri Nov 14 16:45:35 EST 2008

Hi list readers,

The paper mentions the following step:

The problematic coupling between \theta and \beta arises due to the
edges between \theta, z, and w. By dropping these edges and the w
nodes, and endowing the resulting simpliﬁed graphical model with free
variational parameters, we obtain a family of distributions on the
latent variables.

As I am not a native english speaker can someone please tell me what
"endowing" means in this context? Also, why is it possible to remove
the edges and nodes, and create some completely different model, yet
the one approximating things the way we want? I expect this has
something to do with the type of distribution ?

The other question is related to EM algorithm. It was already pointed
to me that parameter estimation should be done first, and then the
inference, but the paper not only explains these in reversed order,
but the EM algorithm outline is specified as follows:

1. (E-step) For each document, ﬁnd the optimizing values of the
variational parameters
This is done as described in the previous section.

2. (M-step) Maximize the resulting lower bound on the log likelihood
with respect to the model parameters \alpha and \beta. This
corresponds to ﬁnding maximum likelihood estimates with expected
sufﬁcient statistics for each document under the approximate posterior
which is computed in the E-step.

So the first step of the algorithm explicitly relies on the inference
phase being completed as the variational parameters are valid
(optimized) only after inference phase (?).

I'm kinda stuck on this one.

I'd appreciate any hints in the right direction.

best,
Mateusz