Topic:
Don’t Expl-AI-n Yourself: Exploring "Healthy" Models in Machine Learning for Health
Thursday, October 15, 2020 - 2:30pm
Speaker:
Prof. Marzyeh Ghassemi (University of Toronto)
Abstract / Description:
Despite the importance of human health, we do not fundamentally understand what it means to be healthy. Health is unlike many recent machine learning success stories - e.g., games or driving - because there are no agreed-upon, well-defined objectives. In this talk, Dr. Marzyeh Ghassemi will discuss the role of machine learning in health, argue that the demand for model interpretability is dangerous, and explain why models used in health settings must also be "healthy". She will focus on a progression of work that encompasses prediction, time series analysis, and representation learning.
Bio:
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 and Canada Research Chair. She will be moving to MIT's EECS/IMES in July 2021. She has served as a NeurIPS 2019/2020 Workshop Co-Chair, and General Chair for the ACM Conference on Health, Inference and Learning (ACM CHIL). Previously, she was a Visiting Researcher with Alphabet's Verily and a post-doc with Dr. Peter Szolovits at MIT. Prior to her PhD in Computer Science at MIT, Dr. Ghassemi 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. Her work has been featured in popular press such as MIT News, NVIDIA, Huffington Post. She was also recently named one of MIT Tech Review’s 35 Innovators Under 35.
Suggested Readings:
- A Review of Challenges and Opportunities in Machine Learning for Health
- Clinical Intervention Prediction and Understanding Using Deep Networks
- Clinically Accurate Chest X-Ray Report Generation
- Can AI Help Reduce Disparities in General Medical and Mental Health Care?
- Hurtful Words: Quantifying Biases in Clinical Contextual Word Embeddings
- ClinicalVis: Supporting Clinical Task-Focused Design Evaluation
