Fairness in Machine Learning
- Algorithms are used to make or assist with many decisions which impact our lives in fields like marketing, medicine, law, or finance. We need to ensure that algorithmic decisions aren't biased in undesirable ways.
- Fair algorithms could be used to prevent illegal discrimination by companies and justice systems, make more accurate predictions about clinical treatments , or provide more diverse content filtering methods on social media.
- I am interested in exploring and defining algorithmic fairness, as well as applying and integrating it with deep learning methods.
- By this same token, AI algorithms that are intended to operate in practice cannot endanger their users or people in general.
- To be safely usable, a model must be robust to reward misspecification and able constructively interact with humans.
- I am interested in thinking about AI safety and transparency and its practical implications on a technical and policy level.
- Learning Adversarially Fair and Transferable Representations
David Madras , Elliot Creager, Toniann Pitassi, Richard Zemel
- Predict Responsibly: Increasing Fairness by Learning to Defer
David Madras , Toniann Pitassi, Richard Zemel
NIPS 2017 Workshop on Transparent & Interpretable ML in Safety Critical Environments (Oral, Best Paper) and NIPS 2017 Interpretable ML Symposium
- Change-point Detection Methods for Body-Worn Video
Stephanie Allen*, David Madras*, Ye Ye*, Greg Zanotti*
SIURO (Vol. 10), August 2017 and Joint Math Meetings (JMM) 2017
(* equal contribution)
I was a TA for the following courses:
- Probabilistic Learning and Reasoning, Winter 2017
- Introduction to Machine Learning, Fall 2016
- Introduction to Computer Science, Winter 2016, Fall 2017
- Introduction to Computer Programming, various semesters 2013-15
Some presentations I've made for various reasons:
- Predict Responsibly: Increasing Fairness by Learning to Defer NIPS 2017 Workshop on Transparent and Interpretable Machine Learning in Safety Critical Environments (Best Paper Award)
- Fairness in Machine Learning: An Overview U of T Undergraduate AI Group (UAIG) AI Day, 2017
I'm also a musician - I love writing, singing, and playing music. A couple of years ago I wrote the songs for a musical in the Toronto Fringe Festival and I'm currently working on some more music, to be released soon-ish (at the time of writing this webpage). I also play jazz piano and love to improvise. If you're interested, check out my Youtube channel!
lastname at cs dot toronto dot edu