Individual Based Modelling in Epidemiology: A Practical Introduction 14-18 May 2018, Stellenbosch, South Africa

Posted on Thu, Dec 14 2017 14:29:00

Prof Wim Delva (SACEMA, Ghent University, Hasselt University and KU Leuven) and Dr Lander Willem (Antwerp University) will be presenting an intensive five-day course on Individual-based modelling in epidemiology, organised by SACEMA. The course, to be registered as a University of Stellenbosch Short Course, will run from 9 am to 4 pm daily, 14-18 May 2018, and be held at STIAS (adjacent to SACEMA) in Stellenbosch. The fee structure and registration form will be available online at sacema.org in early February. This Flyer may be downloaded for printing or circulating.


Course Overview

Individual-based models (IBMs), also frequently referred to as agent-based models, are a relatively new class of models that can be used to gain insight into the population dynamics of complex systems that emerge from the characteristics and behaviours of individuals in the population. This course aims to give participants the skills to design, implement, analyse and calibrate IBMs. First, we will introduce the fundamentals of modelling, as well as a popular, open-access platform for building IBMs (NetLogo). Next, several key modelling concepts will be discussed, including:

  • Deciding on appropriate structure and complexity of the model
  • Developing rules for the actions of individuals
  • Analysing the emerging model dynamics
  • Fitting the model to empirical data
  • Validating model inputs and outputs
  • Communicating the model results

Each of these concepts will be illustrated with practical examples that participants can run on their own laptops. Using the open access software NetLogo and R, tutorials will revolve around model applications to study the epidemiology of HIV, influenza, malaria and diabetes. In the second half of the course, participants will be asked to modify, extend, analyse or calibrate one of these “classroom” models. In addition to strengthening skills in NetLogo and R programming, these mini-projects are also important exercises in interdisciplinary communication for the two- or three-member project teams, as they require teams to agree on a well-defined research question, an appropriate model world, an informative analysis of model output and valid conclusions.

 

Target Audience

  • Post-graduate students and health science professionals whose work potentially involves the design and/or use of IBMs in epidemiology.
  • Ideally, participants should have used R previously for data analysis. However, this is not an absolute prerequisite. R newbies will be asked to work through selected R tutorials before arriving at the course. Prior experience with NetLogo is not required

Lecturers

A medical doctor and epidemiologist, Wim Delva (MD, PhD) has a joint research appointment at SACEMA (Stellenbosch University, South Africa) as well as Hasselt University, Ghent University and KU Leuven in Belgium. He is interested in the application of the statistical, epidemiological and mathematical modelling techniques to describe and analyse the behavioural and biological processes underlying HIV epidemics in sub-Saharan Africa and Europe. His current research centres around the inference of sexual network structure, using phylogenetic tree data and behavioural survey data. Other ongoing research projects seek to explore the role of age-mixing patterns and heritability of HIV set point viral load in HIV transmission dynamics, as well as the impact of biomedical and behavioural interventions on HIV incidence.


 

Holder of an interdisciplinary PhD in Medical Sciences & Sciences, Lander Willem (MSc, PhD) is interested in individual-based modeling with a particular focus on model exploration, parameter estimation, social contact patterns and computational performance. He is currently employed as postdoctoral research fellow at the University of Antwerp.

 

Contact Us

Enquiries that are not already addressed above may be directed to Gavin Hitchcock <aghitchcock@gmail.com> and cc to Wim Delva <Wim.Delva@UGent.be>.