Statistics Masterclass II: Bayesian regression
with Natalia Levshina (Radboud University)
After the success of the first round, the Statistics Masterclass is back!
Many of you have expressed interest in Bayesian statistics, so we are pleased to offer a special opportunity. Natalia Levshina, a statistics expert and author of How to Do Linguistics with R Natalia Levshina (Radboud University), will lead a 4-day masterclass on Bayesian regression in R. This course is ideal for those with a basic understanding of linear (mixed) models and experience using R.
Each day consists of two sessions (morning and afternoon) of two hours each, except for the second day, which will only have a morning session. The morning sessions are more theoretical and accessible online, whereas the afternoon sessions are interactive, practical exercises with R and will require in-person participation. The detailed schedule can be found below.
When? October 21st - 24th
Where? This is a partially hybrid event that will take place both online on Zoom and at the University of Tartu in Ülikooli 18-139 and Jakobi 2-103. You will get the Zoom link after registration.
Schedule (all times in Eastern European time)
Day 1: Monday October 21st (Ülikooli 18-139)
First session (hybrid): 10:00-12:00 EE
Second session (Tartu only): 14:00-16:00 EE
Bayesian inference: concepts, MCMC, comparison with frequentist methods, R packages rstan and brms.
Continuation
Day 2: Tuesday October 22nd (Ülikooli 18-139)
First session (hybrid): 10:00-12:00 EE
TÜling talk (Jakobi 2-438)
Bayesian model selection and diagnostics, the role of priors.
No workshop (instead, Natalia will give a colloquium talk at the University of Tartu)
Day 3: October 23rd (Ülikooli 18-139)
First session (hybrid): 10:00-12:00 EE
Second session (Tartu only): 14:00-16:00 EE
Bayesian Generalized Linear Models: logistic and beta.
Continuation
Day 4: October 24th (Jakobi 2-103)
First session (hybrid): 10:00-12:00 EE
Second session (Tartu only): 14:00-16:00 EE
Visualization, interpretation and reporting Bayesian models.
Discussions of your potential applications for what you've learned.
Preparation?
This masterclass is suitable for everybody who has either participated in the previous Statistics Masterclass, or already has some knowledge of statistics.
About the instructor
Natalia Levshina is Assistant Professor at the Centre for Language Studies at Radboud University. She is the author of How to Do Linguistics with R.
Her research concerns communicative efficiency and its impact on grammar and the lexicon in the long run. She has expertise in working with large corpora, typology, data visualization, machine learning, and frequentist and Bayesian Generalised Linear Mixed Models.
Natalia also documents her linguistic findings about politics, gender and AI in her popular science blog Text Grinder.
