Talk Title: Inference in the Presence of Incompatibility
Endless jokes tell us that Correlation ≠ Causality, from which we can infer that causality needs something more and ask “Correlation+WHAT = Causality?” In philosophy, Causal Modeling introduces ideas of intervention and different kinds of graphical and counterfactual logic tools that sit a little uncomfortably on top of classical probability; in mathematics and physics, Generalized Probability and contextuality give us similar tools in a more unified framework; for machine learning, there’s something completely different.
Peter W. Morgan
Coming out of a mathematics degree and a dozen years as a computer programmer, my research has over the last 30 years become focused on our understanding of the relationship between classical and quantum field theories. Having followed my wife’s career from University College London to Oxford to Yale, I have been a Lab Associate in Yale Physics since 2004.