Presentations of this event are now available:
Interpretable machine learning and causal inference are both hot topics, related in the kinds of problems they can be applied to. Each aims to address deficiencies in conventional machine learning and statistical approaches to model building. We believe researchers and practitioners working in each community have much to learn from each other, but that without first establishing common ground and defining clear boundaries, communication and collaboration will be difficult. In this workshop, we will hear from four experts about their own research investigating methods and applications in these two areas, which we hope will highlight both the commonalities and differences between them. We also anticipate a lively discussion after the presentations.
13:00 - 13:15 Introduction & welcome
13:15 - 13:55 Peter Tennant - "Table 2 Fallacy: Or why interpretation needs more than transparency"
13:55 - 14:00 - Break -
14:00 - 14:40 Noemi Kreif - "Using causal machine learning to explore heterogeneous responses to policies"
14:40 - 15:05 - Break (14:40 - 14:55 Comp Stats & ML Section AGM during this break) -
15:05 - 15:45 Alessandra Russo - "Symbolic machine learning for interpretable AI: recent advancements and future directions"
15:45 - 15:50 - Break -
15:50 - 16:30 Vera Liao - "Questioning the AI: towards human-centered interpretable machine learning"
16:30 - 17:00 Further questions/general discussion/wrap up
Richard Tomsett on behalf of the RSS Computational Statistics and Machine Learning Section
Free to RSS Fellows and e-students
£10 for non-members
Join the Society
as a Fellow.
(for free) as an e-Student.