Distinguished Guest Seminar - Jeff Leek
Distinguished Guest Seminar - Jeff Leek
The DSI Distinguished Guest Seminar will feature Dr. Jeff Leek, Vice President and Chief Data Officer at Fred Hutch. Dr. Leek is a leading expert in biostatistics and data science, with significant contributions to research and data-driven strategies in public health. His talk, titled "Post Prediction Inference – What Do We Do After We Have Machine Learned Everything?", will explore how statistical methods handle machine learning uncertainty and the implications of a data-driven future. The seminar will be available for both in-person and remote viewing, providing an opportunity to gain valuable insights from one of the field's foremost authorities.
Talk Abstract
AI and machine learning are on everyone’s mind! The major advances in generative AI over the last year have inspired the imagination of many people across the biomedical sciences. At Fred Hutch, we have been developing tools and technologies to enable the application of AI both locally and at a national scale through our Cancer AI Alliance. But what happens when we have machine-learned everything? As we move toward a world where more and more of our data is predicted, we need statistical methods that can appropriately handle “machine learning uncertainty.”
In this talk, I will discuss some of the work we have done to build AI infrastructure, how predictions have become a new type of data, and highlight statistical methods developed by our group and others that enable inference with predicted data (IPD). This is joint work with many people, but particularly Awan Afiaz, Anna Neufeld, Kentaro Hoffman, Steve Salerno, and Tyler McCormick.