Learning Bayesian Statistics

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

Apr 16, 2021
Ask episode
AI Snips
Chapters
Transcript
Episode notes
ANECDOTE

Bayesian Breakthrough

  • Sean Taylor's econometrics training limited his options for a complex model.
  • Bayesian methods and Stan helped him create a more flexible model, a breakthrough moment.
INSIGHT

Modeling as Design

  • Modeling is like designing a container for data, offering a flexible approach.
  • Bayesian models allow for a nesting structure and gradual weakening of priors.
INSIGHT

Prophet's Approach

  • Prophet uses curve fitting for forecasting, focusing on business time series data.
  • It offers a control panel of priors, allowing users to inject domain knowledge.
Get the Snipd Podcast app to discover more snips from this episode
Get the app