Bayesian Reasoning
Learn how evidence updates uncertainty and why posterior distributions are more useful than single estimates.
What Learners Build
The course is organized around complete statistical workflows: model design, prior selection, computation, diagnostics, interpretation, and communication.
Computational Modeling
Use simulation, Monte Carlo, MCMC, PyMC, and ArviZ to fit models that do not have closed-form solutions.
Research Communication
Prepare model reports, posterior visualizations, diagnostic summaries, and capstone research outputs.
Curriculum
| Week | Module | Outcome |
|---|---|---|
| 1 | Bayesian Thinking | Understand uncertainty and evidence updating. |
| 2 | Probability Review | Use conditional probability and Bayes theorem. |
| 3 | Priors, Likelihoods, and Posteriors | Translate questions into Bayesian models. |
| 4 | Conjugate Models | Solve beta-binomial, gamma-Poisson, and normal models. |
| 5 | Monte Carlo Computation | Approximate posterior summaries by simulation. |
| 6 | MCMC | Run chains and diagnose convergence. |
| 7 | Hierarchical Models | Model grouped data with partial pooling. |
| 8 | Bayesian Regression | Fit linear, logistic, and count models. |
| 9 | Model Checking | Evaluate predictive fit and compare models. |
| 10 | Decision Analysis | Turn posterior uncertainty into decisions. |
| 11 | Causal and Time-Series Extensions | Extend Bayesian thinking to advanced applications. |
| 12 | Capstone Workflow | Complete a reproducible Bayesian project. |
Start Learning
Begin with the syllabus, then move through the weekly modules and labs. Each assignment is designed to become part of the final capstone portfolio.