SVEPM 2019 workshop on Multivariate analysis using Additive Bayesian Networks

The SVEPM workshop on Multivariate analysis using Additive Bayesian Networks is part of the Conference and annual general meeting (AGM) of the Society for Veterinary Epidemiology and Preventive Medicine (SVEPM2019) from March 27-29th, 2019 in Utrecht (Netherlands).

**When:** March the 29th 2019

**Where:** Utrech the Netherland

**Requirements:** Personal laptop. Basic statistics and basic knowledge of R. No prior knowledge about graph theory or Bayesian statistics is needed. Please follow the getting started checklist

**Instructors:** Arianna Comin & Gilles Kratzer

Analysing animal health data is challenging, as the health status of individuals or groups of animals might depend on many inter-related variables. Such complexity can be underestimated by traditional multivariable regression models, such as linear or generalized linear models, where one variable is designated as a single response variable and all the remaining variables as predictors. Therefore, a holistic, multidimensional approach may be preferable when conducting epidemiological analyses of complex biological data. Additive Bayesian network modelling is a modern computational and graphical approach to analyse complex systems, which extends generalized linear models to multiple dependent variables. The focus is on structure discovery, that means determining an optimal statistical model directly from observed data, allowing all variables to be potentially mutually dependent.

- To understand the basic theory behind structure discovery and additive Bayesian networks
- To learn how to set up and interpret an additive Bayesian network model for multivariate analysis of animal health data using the R package abn

Time | Topic | Slides | R code |
---|---|---|---|

8:00 - 8:30 | Bayesian Networks in a Nutshel | ||