Course content

Principles of statistical modeling and inference. Likelihood theory. General theory for generalized linear models, with applications to regression models for normally distributed data, logistic regression for binary and multinomial data, Poisson regression models and log-linear models for contingency tables. Extensions of GLM-theory to, for example, models for over-dispersion and quasi-likelihood estimation.

Learning outcome

The course gives an introduction to generalized linear models (GLM), which is a natural generalization of ordinary (multivariate) linear regression for normally distributed responses to allowing responses from a broader class of distributions, in particular discrete distributions. The students will through the mandatory exercises be capable of using the theory to analyze data sets.

Learning methods and activities

Lectures, exercises and a project/term paper with the use of computer (statistical package R). The lectures may be given in English. Portfolio assessment is the basis for the grade awarded in the course. This portfolio comprises a written final examination 70% and the project/term paper 30%. The results for the constituent parts are to be given in %-points, while the grade for the whole portfolio (course grade) is given by the letter grading system. Retake of examination may be given as an oral examination.

Compulsory assignments

• Exercises

Recommended previous knowledge

TMA4267 Linear Statistical Models or TMA4255 Applied Statistics.

Course materials

Will be announced at the beginning of the semester.