Bayesian Time Series
summary
Practical checklist: tutorial uses PyMC in Jupyter (Conda / Pixie setups) to build AR/AR1/AR2, random‑walk and state‑space models in PyMC, add Fourier seasonality and holiday effects, and fit Gaussian process priors; examples run on US daily births (228 months spanning 1970–1988) and Arctic ground squirrel densities where the population later collapses to zero in 2009. It shows actual outcomes: MCMC sampling with posterior predictive checks and diagnostics (R_hat, ESS, divergences), when forecasts succeed and when they fail (regime change like a predator pit), and how models and priors are iteratively adjusted.
AI-written summary · about
chapters (9)
- 0:00Bayesian time series intro
- 3:16Time series analysis importance
- 14:02Bayesian vs frequentist basics
- 28:05First model with PMC
- 41:12Ground squirrel population modeling
- 59:19Advanced time series models
- 1:01:20Holiday effects in births model
- 1:18:28Gaussian process models intro
- 1:25:03Advanced resources & next steps
related talks
01h 27mProbabilistic Python: An Introduction to Bayesian Modeling with PyMC
42mBayesian Non-parametric Models for Data Science using PyMC3
01h 26mProbabilistic Programming and Bayesian Computing with PyMC
42mSolving Real-World Business Problems with Bayesian Modeling
44mAn introduction to PyMC3
01h 31mProbabilistic Programming and Bayesian Inference in Python
34mFast, beautiful and easy Bayesian modelling: Can you have it all?
01h 40mApplied Time Series Econometrics in Python and R