1
$\begingroup$

I am planning to present a simple introduction to variational Bayesian inference during a lab research meet (45 mins). My audience is mainly from a psychometrics, psychology, educational sciences background who have a basic idea of Bayesian inference and mainly work with SPSS. Computational methods are relatively rare here and not discussed.

I was wondering how do I structure the presentation, I would lose the audience if I make it math heavy, yet, it is hard to talk about variational methods hiding the math.

I do not want to present it as a look, a cool method, but more to make people aware that there are alternatives to MCMC if they want to do Bayesian inference (in a large dataset and complex model context). Any suggestions would be highly appreciated!

$\endgroup$
1
$\begingroup$

I was in a similar position a couple years ago and stumbled upon this. The author has lifted a lot of the material straight from Bishop (ch. 10) but oh well. What I thought was quite good for at least the first portion was it serving as a decent look at the topic from a heuristic point of view (pros and cons) with some math present as well. I found it a good source of motivation.

I'd share what I actually did, but alas I don't have the presentation any more. Good luck!

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.