Doktorandseminarium: Bayesian model selection and statistical control in the social sciences using Gaussian processes
- Location: Ångströmlaboratoriet, Lägerhyddsvägen 1 4007
- Lecturer: Björn Blomqvist
- Contact person: Volodymyr Mazorchuk
Gaussian Processes is a generic method for supervised learning often used for regression and classification. In this talk, I will show how this methodology can be used to study social systems. We will see that Gaussian processes can formulate a Bayesian framework for model selection in regression and that we, by our choices of mean and covariance functions, can retain both interpretability and predictability in our models by allowing flexible nonlinear statistical controls. I will show how we can use this method with an example - the rise of the Swedish right-wing populist party, Sverigedemokraterna.