Course content

The course is given each second year, next time autumn 2025. The course will cover: Review of statistical methods. Mathematical models: Empirical models Models based cause and effect relations. Model fitting Linear models Non linear models Model discrimination Design of experiments Response surface design Design for non-linear models Compulsory computer exercises and projects are part of the course.

Learning outcome

Knowledge: The students will get knowledge about two different paths to modelling, empirical and physical. They will obtain knowledge about applied numerical methods, such as steepest descend, Newton iteration, numerical integration of ordinary differential equations, in addition to optimization methods. Skills: By completing the couse, the students are able to do linear and non-linear regression. They will be able to fit model parameters to measurements and estimate the confidence interval of parameters and model predictions. The students will be able to fit multi-response models to experimental data. They will be able to develop dynamic and steady-state models that will be used for explaining experimantal data. They will be able to do regression when there measurement errors in all variables, both in design and response variables. The students should be able to apply methods for model discrimination. Moreover, they should be able to apply methods for doing experimental design. General competence: The students will gain competance in using tools for model fitting and programming in matlab and/or python.

Compulsory assignments

• Project 1
• Project 2

Further on evaluation

2 projects need to be approved in order to take the exam.

Required previous knowledge

Elementary knowledge in statistics, numerical methods, linear algebra and computer programming.

Course materials

Handouts

Credit reductions