Time Series Analysis with Bayesian State Space Models in PyMC | Jesse Grabowski | PyMC Labs

📒 Presentation Notebooks: https://github.com/jessegrabowski/sta...

Time series are everywhere, and building time into our models can bring them to the next level. Modeling time series, however, can be a minefield. They can be especially hard in PyMC, where one needs to have deep knowledge of both pytensor and PyMC internals to set up recursive models, handle missing data, make out of sample forecasts, or generate IRFs. This talk aims to introduce PyMC users to the pymc-statespace module, a collection of tools designed to help users past these hurdles and into time series modeling. The talk will introduce the linear gaussian state space framework, and give end-to-end bayesian time series workflow examples. This will include pre-defined models like SARIMAX, modular structural time series models built from latent components, and an example of implementing a complex custom model, using a Dynamic Stochastic General Equilibrium (DSGE) as a motivating example.

Resources
Example notebooks in pymc-experimental:
SARMA: https://github.com/pymc-devs/pymc-exp...
VARMA: https://github.com/pymc-devs/pymc-exp...
Structural Modeling: https://github.com/pymc-devs/pymc-exp...
Making a custom state-space model: https://github.com/pymc-devs/pymc-exp...

💼 About the speaker:
Jesse Grabowski
PhD Candidate, Paris 1 Pantheon Sorbonne
Principal Data Scientist, PyMC Labs
Jesse Grabowski is an economist and data scientist, specializing in macroeconomics, finance, and machine learning. He has worked as a economic modeler at the Organization for Economic Cooperation and Development (OECD) Development Centre for several years. He currently works as a data scientist at PyMC labs, where he helps deliver elite-level solutions to complex business problems. He is currently a PhD candidate at Paris 1 Pantheon-Sorbonne, where his work focuses on quantifying the macroeconomic risks of tail climate risks using Bayesian methods.
🔗 Connect with Jesse:
👉 Linkedin:   / jessegrabowski  
👉 Github: https://github.com/jessegrabowski
👉 Youtube:    / @thayme  
👉 Website: http://jbgrabowski.com/

💼 About the Host:
Thomas Wiecki (Founder of PyMC Labs)
Dr. Thomas Wiecki is an author of PyMC, the leading platform for statistical data science. To help businesses solve some of their trickiest data science problems, he assembled a world-class team of Bayesian modelers and founded PyMC Labs -- the Bayesian consultancy. He did his PhD at Brown University studying cognitive neuroscience.
🔗 Connect with Thomas:
👉 Linkedin:   / twiecki  
👉 GitHub: https://github.com/twiecki
👉 Twitter:   / twiecki  
👉 Website: https://www.pymc-labs.com/
https://twiecki.io/

🔗 Connecting with PyMC Labs:
🌐 Website: https://www.pymc-labs.com/
👥 LinkedIn:   / pymc-labs  
🐦 Twitter:   / pymc_labs  
🎥 YouTube:    / pymclabs  
🤝 Meetup: https://www.meetup.com/pymc-labs-onli...