Introduction to prediction modelling
This short course provides a practical introduction to prediction modelling over two half days: Monday, November 9-Tuesday, November 10 from 1 PM-5 PM at the School of Population and Global Health, University of Melbourne. There is also an option for zoom attendance, although in person attendance is preferred.
The course starts with an overview of the main types of research questions in which data analysis plays a central role, and how these differ from traditional hypothesis-driven studies. It then explains the key concepts and tools needed to develop diagnostic and prognostic prediction models, assess predictive model performance, and carry out internal and external validation. These approaches are essentially different from hypothesis testing, p-values and confidence intervals as used in causal models.
The course covers both statistical regression models and machine learning approaches, illustrated with case study examples. Teaching includes both classroom lectures and computer practicals, with a focus on the application of diagnostic and prediction models.
Who should attend?
This course is suitable for graduate research students, early-career researchers, and health and clinical researchers who are interested in developing or strengthening their understanding of how to develop, validate and interpret prediction models.
Prerequisites
Participants should be familiar with either R or Stata, for example being able to create new variables, run regression commands and obtain basic plots.