pymc-bayesian-modeling

What problem does it solve?

This Skill introduces PyMC for Bayesian modeling and probabilistic programming. It enables building, fitting, validating, and comparing Bayesian models (linear, hierarchical, time series) with MCMC (NUTS) and variational inference, plus posterior checks and model comparison.

Core Features & Use Cases

• End-to-end Bayesian workflow: Build priors, define likelihoods, sample posteriors, and validate with posterior predictive checks.
• Model comparison & diagnostics: Use LOO/WAIC, R-hat, ESS, and divergences to select robust models.
• Hierarchical modeling: Implement multilevel structures with non-centered parameterization for better convergence.

Quick Start

Run a simple Bayesian linear regression, check diagnostics, and compare two models using LOO.