Introduction to Generalized Nonlinear Models

Users from all professional groups who have basic knowledge in statistics and have used R before. (For introductory R courses please revisit the course list.)

• CHF 400 for members of UZH/ETH and associated institutes
• CHF 500 for members of other universities, federal and cantonal research facilities and agencies, non-profit organisations etc.
• CHF 800 for companies, banks, insurance companies etc.

The class of generalized linear models encompasses many tools commonly used in data analysis, including multiple linear regression, logistic regression, log-linear models, etc. But a linear predictor does not always capture the relationship we wish to model. Rather, a nonlinear predictor may provide a better description of the observed data, often with fewer and more interpretable parameters.

This workshop introduces the wider class of generalized nonlinear models and their implementation via the R package gnm. The day will begin with a brief refresher of linear and generalized linear models. The extension to a nonlinear predictor will be motivated by the case of structured interaction models, such as Goodman's Row-Column association model for contingency tables. The remainder of the day will be devoted to further applications, such as the UNIDIFF models for social mobility data, the diagonal reference model for dependence on a square classification and the Lee-Carter model for mortality data.

For all Zurich R Courses participants should bring their own laptops to the course and will be informed by email in advance which packages they need to install.