Distributional Regression with applications in Biomedicine, Biology, and Economics
Santiago de Compostela. February 24-26, 2015
In many modern applications, one is not only interested in explaining the effect of covariates on the expected outcome but would rather try to regress the complete distributional on explanatory variables.
In this course, we will introduce distributional regression, a generic framework for performing regression analyses where several parameters of a potentially multivariate response distribution are related to flexible regression predictors.
Although classic regression analyses entail easy interpretation, they only focus on means and averages and therefore may lead to erroneous conclusions when modeling complex data structures. The distributional regression framework allows us to overcome these problems.
To fully exploit the capabilities of distributional regression, we will consider structured additive regression specifications, which allow for combination of nonlinear effects of continuous covariates, spatial effects, random effects and a number of extensions.
The Distributional Regression methodology can be used in a variety of disciplines including: medicine, biology, genomics, ecology, marine research, and economics and finance.
Statistical software: R
Main R packages: BayesR, R2BayesX, and gamlss.
Statisticians and researchers in the areas of biomedicine, biology, economics, and other disciplines, working with complex data structures.
18 hours (6 hours per day)
Days: February 24th, 25th and 26th
Timetable: 10:00-14:30 and 16:30-18:30
Thomas Kneib - Georg-August-University Göttingen (Germany).
Nadja Klein - Georg-August-University Göttingen (Germany).
- Fill in the online registration form before February 12th, 2015.
- After registration you will receive further instructions via email.
Limited number of spaces available
You can download more information about this event through the following link(s):