Seminar Series: fAIrest of Them All by Marzyeh Ghassemi, PhD
Seminar Series: fAIrest of Them All
Join us on Zoom
While clinical AI and medical risk scores have received much attention for their potential to achieve above-human performance, there are many concerns about their ability to mimic societal bias. In this talk, Dr. Ghassemi explores the difficulty of making state-of-the-art machine learning models behave as we say, not as we do, and how technical choices that seems natural in other settings may not work well in health.
Dr. Marzyeh Ghassemi is an Assistant Professor at the University of Toronto in Computer Science and Medicine, and a Vector Institute faculty member holding a Canadian CIFAR AI Chair. She currently serves as a NeurIPS 2019 Workshop Co-Chair, and Board Member of the Machine Learning for Health Unconference. Previously, she was a Visiting Researcher with Alphabet’s Verily and a post-doc with Dr. Peter Szolovits at MIT (CV).
Professor Ghassemi’s PhD research at MIT focused on creating and applying machine learning algorithms towards improved prediction and stratification of relevant human risks with clinical collaborations at Beth Israel Deaconess Medical Center and Massachusetts General Hospital, encompassing unsupervised learning, supervised learning, and structured prediction. Her work has been applied to estimating the physiological state of patients during critical illnesses, modeling the need for a clinical intervention, and diagnosing phonotraumatic voice disorders from wearable sensor data. Prior to MIT, Marzyeh received an MSc. degree in biomedical engineering from Oxford University as a Marshall Scholar, and B.S. degrees in computer science and electrical engineering as a Goldwater Scholar at New Mexico State University.