AIMI Journal Club: Improving the Accuracy of Medical Diagnosis With Causal Machine Learning - Jonathan Richens, PhD
The live event is for the Stanford community. The recorded presentation is available here for everyone to view.
Machine learning promises to revolutionize clinical decision making and diagnosis. It has long been known that causal reasoning is fundamental to how doctors perform diagnosis - doctors aim to explain a patient’s symptoms by determining the diseases causing them. However, existing data-driven machine learning approaches fail to differentiate correlation from causation. In this talk, we will explore how this failure to disentangle correlation from causation can lead to suboptimal or even dangerous diagnoses, and how these problems can be resolved by designing algorithms that can perform causal and counterfactual reasoning.
Dr. Jonathan Richens is a senior researcher at Babylon Health. He obtained his PhD from Imperial College London in quantum information theory before moving into machine learning research. Jonathan is currently working on applying machine learning and causal inference to open problems in clinical decision making and algorithmic fairness.
Link to paper