PyMC3 is great for inferring parameter values in a model given some observations, but sometimes we also want to generate random samples from the model as predictions given what we already inferred from the observed data. This kind of sampling is called posterior predictive sampling, and it can be very hard. The typical problems that show up are related to shape mismatches in hierarchical models, latent categorical values that aren’t correctly re-sampled or changing the shape of the data between the training and test phases. In this presentation I’ll talk about how posterior predictive sampling is implemented in PyMC3, show some typical situations where it fails, and how to make it work.
I got into Bayesian stats during my PhD in cognitive neuroscience. During my postdoc I got more involved with machine learning, and discovered PyMC3. I became a core contributor of PyMC, learnt a lot in the process and made up my mind to pursue a career outside of academia. I am now a machine learning engineer at Innova SpA in Italy.
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.