Learning Bayesian Statistics cover image

#37 Prophet, Time Series & Causal Inference, with Sean Taylor

Learning Bayesian Statistics

00:00

Model Checking and Validation for Forecasting

The strength ar empirically, the model works well and when we've seen it give reasonable behavior across a wide variety of data sets. The weaknesses are, ithink installation is definitely one of them. We trid to build it in a way that was extensible to a number of different problems. But i think it's worth acknowledging that it's not a one size fits all solution,. When your time series isn't sort of generated by similar family of processes to that we've seen before, then it's likely to be a poor model. If i could do it all over again, i think that the model checking and model validation would be more at the forefront of the package.

Transcript
Play full episode

The AI-powered Podcast Player

Save insights by tapping your headphones, chat with episodes, discover the best highlights - and more!
App store bannerPlay store banner
Get the app