Authors: Petr Keil, Jan Smyčka
This post contains materials for Bayesian stats course (2-4 Feb 2015 at Faculty of Science, Charles University, Prague, Czech Republic).
The complete materials and their source codes (Markdown and R) are on my GitHub repository. The lectures can also be accessed directly as follows:
- Introduction: Course contents, pros and cons of Bayes, necessary skills.
- Normal distribution: Introducing likelihood and deviance on the Normal example.
- Poisson distribution: The didactic simplicity of Poisson and its likelihood. Likelihood maximization. Probability mass function.
- How to calculate posterior probability (by Jan Smyčka): The principle of MCMC sampling.
- Bayesian resources: Overview of software, books and on-line resources.
- First real model in JAGS: Fitting Poisson distribution to forest larvae count data.
- T-test: First model with 'effects', hypotheses testing, derived variables.
- Linear regression: Extracting credible and prediction intervals.
- Time series analysis: Fitting more complex function. Auto-regression.
- ANOVA part 1 and part 2: Fixed effects, random effects, the effect of small sample, and shrinkage.
- Site occupancy model: accounting for imperfect detection.
- Other probability distributions (binomial, beta, gamma, multivariate normal, negative binomial).
- Convergence diagnostics, publishing Bayesian papers.
- Concluding remarks.