Stan: why does it exist? when is it useful? why do I use it?
summary
Stan exists as a statically typed modeling language plus fast inference and interfaces (C++ autodiff core, HMC sampling, ADVI) to let users encode any differentiable joint probability and avoid ML degeneracies like the hierarchical logistic where maximum likelihood collapses the state variance to zero. It proves useful for compact but complex models with hierarchy and censoring, for example a semi‑mechanistic tumor‑size + survival model that projects 30→300 projection results and supports mc-stan.org workflows and hierarchical logistic example.
AI-written summary · about
chapters (6)
- 0:00Introduction
- 0:47Why Stan exists
- 6:03Stan language and interfaces
- 12:07When Stan is useful
- 16:01Why I use Stan
- 23:36Q&A
