Demystifying Variational Inference by Sayam Kumar

Talk Abstract

What will you do if MCMC is taking too long to sample? Also what if the dataset is huge? Is there any other cost-effective method for finding the posterior that can save us and potentially produce similar results? Well, you have come to the right place. In this talk, I will explain the intuition and maths behind Variational Inference, the algorithms capturing the amount of correlation, out of the box implementations that we can use, and ultimately diagnosing the model to fit our use case.


Sayam Kumar

Sayam Kumar is a Computer Science undergraduate student at IIIT Sri City, India. He loves to travel and study maths in his free time. He also finds Bayesian statistics super awesome. He was a Google Summer of Code student with NumFOCUS community and contributed towards adding Variational Inference methods to PyMC4.

This is a PyMCon 2020 talk

Learn more about PyMCon!

PyMCon is an asynchronous-first virtual conference for the Bayesian community.

We have posted all the talks here in Discourse on October 24th, one week before the live PyMCon session for everyone to see and discuss at their own pace.

If you are available on October 31st you can register for the live session here!, but if you are not don’t worry, all the talks are already available here on Discourse (keynotes will be posted after the conference) and you can network here on Discourse and on our Zulip.

We value the participation of each member of the PyMC community and want all attendees to have an enjoyable and fulfilling experience. Accordingly, all attendees are expected to show respect and courtesy to other attendees throughout the conference and at all conference events. Everyone taking part in PyMCon activities must abide by the PyMCon Code of Conduct. You can report any incident through this from.

If you want to support PyMCon and the PyMC community but you can’t attend the live session, consider donating to PyMC

Do you have suggestions to improve PyMCon? We have an anonymous suggestion box waiting for you

Have you enjoyed PyMCon? Please fill our PyMCon attendee survey. It is open to both async PyMCon attendees and people taking part in the live session.


Nice talk! Now that @Sayam753 you are familiar with both pymc3 and pymc4/tfp kind of VI API, what do you think we should improve on the pymc3 side moving forward?

1 Like

Thanks @junpenglao for the follow up. I would love to see generative modelling through VI going forward in PyMC3. Especially the recent advances in hierarchical variational bayes, NVAE paper. Also, some benchmarks using VI on JAX backend would be great.


Thanks for the talk and the public notebook! :ok_hand:

1 Like