May 11, midday: Made a list of all the things I needed to do to finish the thesis.
May 12, midday: Tear up the list, decide to finish it with what I had at hand.
May 13, 7am: Thesis finished.
It was very satisfying. Right now, I remain on a high from having implemented EP.
> ‘Arrakis teaches the attitude of the knife — chopping off what’s incomplete and saying: “Now it’s complete because it’s ended here.”‘
Indeed.
Yup, I more or less did the same thing.
What is EP?
I had the same question. My best guess is Expectation Propagation. Even if not, it led off on a productive afternoon learning about non-MCMC approaches to approximate the posterior. Would love to see an example of how to implement….
Nick:
An example of EP implemented with all the details is in the new BDA!
Will I find any DAGs or Bayesian networks in BDA?
Only implicitly, in that our models can be viewed in that way.
This is what I find odd about the recurring DAGs vs. hierarchical models debate. They’re basically the same thing.
I guess some edges in a DAG could be deterministic, but as also has been discussed here, a deterministic relationship just a stochastic relationship with a delta function for a distribution.
Excellent! I am glad to hear it, and also to hear it will be working in STAN.