Statistical modelling: Theory and practice

• Formulate and apply common statistical models
• Perform model check and be able to evaluate the adequacy of statistical methods in specific applications
• Be able to interpret and present results from statistical analysis
• Apply estimation methods, in particular the principle of Maximum Likelihood
• Perform likelihood ratio tests, Wald test and use the profile likelihood function to assess parameter unceartainty
• Use simulation based methods for unceartainty estimation
• Be able to perform standard statistical analysis using the statistical software R
• Recognize the possible use and potential of more advanced statistical methods
• Describe and apply generalized linear models
• Use non-parametric (Kaplan-Meier) and parametric methods to describe and compare survial data
• Model time-series data using hidden Markov models

The most commonly used statistical methods will be taught, usually based on practical problems and using the statistical software packages R. In addition, an introduction to more theoretical topics is provided. Emphasis is placed on training and discussing statistical thinking, while introducing the topics that are covered in more depth in the other advanced statistics courses at the department.

• 36 – 49 (Tues 13-17)

02402/02403, Knowledge of statistical methods equivalent to course 02402/02403. Some additional knowledge of probability theory (02405) is an advantage.

Lectures and excercises (integrated) plus solution of a practical assignments.

The course is a link between the introductory course and more advanced course (e.g. 02429 og 02426) offered at the Department.