Building a simple DLM in Pyro
I recently wrote a simple Dynamic Linear Model (DLM) tutorial for Pyro. Although there is room to improve the inference, the behavior of incorporating latent variables with a time-based prior sheds some light that calibration of models with latent prior derived from other casual procedure such as experiment is possible.
Without prior,

With prior,

Enjoy Reading This Article?
Here are some more articles you might like to read next: