Skip to content

study-groups/bayesian-study-group

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

12 Commits
 
 
 
 
 
 
 
 

Repository files navigation

2022 Study Group

Winter 2022 focus is on Richard McElreath 2022 course on Bayesian statistics. Pdf of first two chapters of his Statistical_Rethinkng gives foundational philosophy. Course ends on the subject of state-space representation of time-series data. Calendar is from McElreath's site.

Calendar & Topical Outline

There are 10 weeks of instruction. Links to lecture recordings will appear in this table. Weekly problem sets are assigned on Fridays and due the next Friday, when we discuss the solutions in the weekly online meeting.

Lecture playlist on Youtube: <Statistical Rethinking 2022>

Week ## Meeting date Reading Lectures
Week 01 07 January Chapters 1, 2 and 3 The Golem of Prague, Bayesian Inference
Week 02 14 January Chapter 4 Basic Regression, Categories & Curves
Week 03 21 January Chapters 5 and 6 Confounding, Even Worse Confounding
Week 04 28 January Chapters 7 and 8 Overfitting, Interactions
Week 05 04 February Chapters 9, 10 and 11 Markov chain Monte Carlo, Binomial GLMs
Week 06 11 February Chapters 11 and 12 Poisson GLMs, Ordered Categories
Week 07 18 February Chapter 13 Multilevel Models, Multi-Multilevel Models
Week 08 25 February Chapter 14 Varying Slopes, Gaussian Processes
Week 09 04 March Chapter 15 Measurement Error, Missing Data
Week 10 11 March Chapters 16 and 17 Beyond GLMs: State-space Models, ODEs, Horoscopes

History of Bayesian statistics

Devised around in 1763 by Thomas Bayes, Bayesian Statistics addresses conditional probability in terms of the ratio between an event A happening simultaneously with another event, B written: P(A given B) = P( A and B ) / P( only A ).

Bayes described conditional probability- not how to update priors or measure probability beyond counting.

In 1814, Laplace presents the classical interpretation of probability which treats probability as a measure of a ratio between a favorable interpretation of an event and all those events not deeemed favorable. In this way, more experimental data leads to tighter confidence intervals — e.g. 90% of the experiments (confidence level) fall between θlow and θhigh (confidence interval). This is the language of classic statistical hypothesis testing.

In 1857 logicians Venn and Boole introduce the frequentist interpretation of probability which treats probability as the measure of likelihood after attempting a large number of trials.

Primary Resources

About

A collection of Bayesian inspired concepts.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published