Causal learning goes beyond standard descriptive or predictive data modelling by allowing researchers to ask what would happen under a change to a system – such as the use of a treatment (in medicine), a change of policy (in economics or government) or a change in product or system attributes (in business). Questions of this kind are of crucial importance in practice, but involve causal notions that can go beyond the theoretical frameworks underlying standard data science tools. Recent advances spanning multiple fields (including machine learning, statistics, econometrics, philosophy and more) are delivering new approaches that have the potential to leverage large datasets and emerging experimental designs to build truly causal models at unprecedented scale.

The Cambridge Centre for Causality brings together world-leading expertise across multiple disciplines to advance the boundaries of causal data science, spanning research in theory and foundations, methods and computing and challenging applications, including in biomedicine and health research. The vision is to leverage and build upon recent advances in our understanding of causality to enable robust and scalable analysis going beyond established descriptive or predictive paradigms.