The AI Health Podcast

Interpretability and Data Challenges with Duke University's Dr. Cynthia Rudin

Episode Summary

Our guest today is Dr. Cynthia Rudin, a professor of computer science, electrical and computer engineering, and statistical science at Duke University. She describes her work on data-driven risk prediction models, which have been validated in real Intensive Care Units and provide a simpler, effective alternative to neural nets. She also discusses the benefits of model interpretability and the challenges that researchers face when accessing medical data.

Episode Notes

Our guest today is Dr. Cynthia Rudin, a professor of computer science, electrical and computer engineering, and statistical science at Duke University. She describes her work on data-driven risk prediction models, which have been validated in real Intensive Care Units and provide a simpler, effective alternative to neural nets. She also discusses the benefits of model interpretability and the challenges that researchers face when accessing medical data.

Pranav and Adriel first provide context for the interview, giving an overview of prognostic models, interpretability and black box models, and GDPR and CCPA, which starts at 13:05. If you like what you hear, let a friend know, subscribe wherever you get your podcasts, and connect with us on Twitter @AIHealthPodcast.