Bayesian approaches in medical diagnosis: principles, tools, and applications.
Lecturer: Luzia Gonçalves | GHTM | IHMT-NOVA
Despite the increasing popularity of Bayesian statistics, some concepts (e.g., prior information, posterior distribution, credibility intervals) remain unclear for non-statisticians. The evaluation of diagnostic tests is a good example to explain Bayesian statistics. In this work, Bayesian latent class models will be discussed to estimate sensitivities and specificities using multiple imperfect diagnostic tests.
No mathematical knowledge is required for this talk.