[MUSIC] So congratulations, you guys have made
it to the end of the last module of this course and so we've covered lots and
lots of material in this course and in particular the last couple of modules have
been quite advanced and quite intense. So in this module, in particular,
we covered Latent Dirichlet Allocation, or LDA, and Gibbs sampling. So LDA is a really widely used tool for
mixed membership modeling and text corpora, but variance of
the LDA model could be used for mixed membership modeling and
a huge range of different applications. And we also covered Gibbs sampling, which
is the most widely used algorithm for Bayesian inference. We specifically examined it
in the context of LDA, but it tends to be the most straightforward
algorithm to think about deriving the updates for
in any Bayesian model. That said, it's not always
the most scalable algorithm, especially just the first implementation
you might think of writing down, but there's a lot of work in a community. I'm thinking about scaling
up Gibbs sampling or Gibbs sampling type algorithms to really
large data sets and really big models. So in conclusion, I want to leave you with
a list of things that you should be able to do now that you've
completed this module. So take a little bit of time
to reflect on this list and think about all the advances
that you've made. And finally, I want to give
a big thanks to David Mimno for providing an outline of the example
that we walked through when we talked about collapse Gibbs sampling and LDA. [MUSIC]