Causal machine learning in health research 

Abstract:

Causal inference is a central goal of health research, where we are interested in understanding the impact of a treatment, policy or other exposures on patient outcomes and population health. For many years, randomised trials were considered the only avenue to undertake causal inference, although recent decades have seen a rise in using observational (non-randomised) data to answer causal questions. However, the need for appropriate design and analytic methods to allow for causal inference from observational data is imperative, particularly in the current era of data deluge, with many developments in this area, including the use of machine learning.
This talk will first provide an overview of some of the fundamental principles of causal inference, before describing the role of machine learning in this space and touch on some ongoing areas of research in this field.

Speaker bio:

Dr Shepherd is a postdoctoral biostatistician and Galli Senior Research Fellow based at The University of Melbourne and Murdoch Children's Research Institute (MCRI). Daisy obtained her PhD at the University of Auckland, before accepting her current position in 2019 within the renowned Victorian Centre for Biostatistics (ViCBiostat). In this role, she conducts biostatistical methods research alongside collaborative statistical work on observational studies in child and adolescent health, working in partnership with the leading state-wide study Generation Victoria (GenV) and the Neurodisability and Rehabilitation group at the MCRI. In her methodological research, she develops and evaluates statistical methods in modern approaches to causal inference, with a focus on causal machine learning and providing practical guidance for researchers.