I make some concluding remarks. Included for completeness, though it's unlikely any YouTube folks will be sitting the exam... ... https://www.youtube.com/watch?v=t3ISn5T2g_I
This video was originally published on lbry.io.
In it, I demonstrate how to install and run DNest4, the software package for Diffusive Nested Sampling (a Markov Chain Monte Carlo algorithm for Bayesian inference, statistics, and related things). For more details, see the paper: https://www.jstatsoft.org/article/view/v086i07
Music: www.bensound.com
...
https://www.youtube.com/watch?v=QkDmd0prS7M
We spotted this flat dead lizard being carried by a solitary ant along the road at Siding Spring Observatory. It was a bit dark so the ant is hard to see but the ant was about 1cm long and the lizard was about 10cm long.
...
https://www.youtube.com/watch?v=ZIz-B5ds8sU
I discuss linear regression and the connection between least squares and Bayesian inference. Then I do Bayesian linear regression using JAGS, and also talk about how to do predictions using JAGS.
...
https://www.youtube.com/watch?v=2zK2sh-_5yU
I discuss Jaynes's famous example demonstrating the conceptual difference between Bayesian credible intervals and classical confidence intervals. I then discuss problems with multiple unknown parameters and why MCMC is useful.
Note: in the R code for verifying the confidence interval, I kept saying that the equals sign should have been +. It actually should have been -. Sorry!
...
https://www.youtube.com/watch?v=E6Uxn95L-FQ
I discuss formal and informal methods of summarising posterior distributions. Point estimates (mean, median, mode) and a bit of decision theoretic justification for them, and credible intervals.
...
https://www.youtube.com/watch?v=TXHYygoPsp4
I discuss a model that can be used in situations like classical one-way ANOVA. It is a hierarchical model. I also discuss (informal) posterior predictive checking.
...
https://www.youtube.com/watch?v=XDNa6K_MSvo
The first lecture of STATS 331 at the University of Auckland. In this lecture I outline the structure of the course and try to provide some motivation for why this subject is worth studying.
RELATED STUFF:
* All the lectures in a single video: lbry://@BrendonBrewer#3/stats331#5
* Supplementary files such as assignments and solutions: lbry://@BrendonBrewer#3/stats331-files#d
* PDF coursebook: lbry://@BrendonBrewer#3/stats331-book#4
I motivate the classical chi-squared test using a game show example, and then develop a Bayesian alternative to it. This involves the Dirichlet and Multinomial distributions, and then a hierarchical model.
...
https://www.youtube.com/watch?v=7Lbx9q6q8AE