Simple inference question

Questions
1

Hi, Im new to PyMC3 and have a simple question about inference of the randomwalk function.
In the Getting Started With PyMC3 docs, there is an example of inference in volatility, but how is this useful? I mean, it is only inferencing what we already know, how about future predictions? From what I understand, only nu, sigma and s are beiing estimated.

image

2

Hey, BjornHartmann. As far as I know, the stochastic volatility model is a state-space model (that focuses on estimating sigma), and as such is a time-series model, which can extrapolate variables (even if they’re sigma) into the future. I recommend you read around more about how time-series models work.