Additive Bayesian Networks as an alternative for risk factor studies

Course description and objectives:

Additive Bayesian Networks (ABN) have been developed to disentangle complex relationship of highly correlated datasets as frequently encountered in risk factor analysis studies. ABN is an efficient approach to sort out direct and indirect relationships among variables which is surprisingly common in systemic epidemiology. The participants are able to run the particular steps within an ABN analysis with real world data. They are able to contrast this approach with standard regression (linear, logistic, Poisson and multinomial models) used for classical risk factor analysis. The participants will be introduced to general machine learning techniques (bootstrapping, variable importance plot, random forest).

Despite the fact that the theoretical foundations of ABN have been well established for decades, the practical implementation is not yet readily available in standard software. The purpose of the organisers is to make this methodology more accessible for wider audience. Ultimately, this workshop aims at empowering residents with new innovative statistical methods relevant for quantitative research.

Lecturers: Sonja Hartnack and Gilles Kratzer

Contact: Sonja Hartnack
PD Dr.med.vet, Dipl. ECVPH
Vetsuisse Faculty
Winterthurerstrasse 270
8057 Zurich
tel +41 44 635 90 51
email: sonja.hartnack@access.uzh.ch

Additional information on registration, venue and suggestions regarding accomodation to be published in due course!