Maggie Makar
About me
I study machine learning, causal inference, and the intersection between the two. My work focuses on building data-efficient causal inference methods in resource-constrained settings, and building robust predictive ML models using ideas from causality. Prior to UM, I completed my PhD in CSAIL, MIT and my bachelors degree in math and economics at Umass Amherst.
Joining my group
I am currently looking for PhD students to work with me. To get a better sense of my work and what my previous student collaborators have achieved please see the students page and the publications page.
Please read how to join before emailing me.
Selected Publications
Full list of publications here
Causally motivated multi-shortcut identification and removal
J. Zheng, M. Makar
NeurIPS, 2022 [workshop version, full paper forthcoming]
Causally-motivated shortcut removal using auxiliary labels
M. Makar, B. Packer, D. Moldovan, D. Blalock, Y. Halpern, A. D'Amour
AISTATS, 2022 [paper]
Exploiting structured data for learning contagious diseases under incomplete testing
M. Makar, L. West, D. Hooper, E. Horvitz, E. Shenoy, J. Guttag
ICML, 2021, [paper]
Estimation of Bounds on Potential Outcomes for Decision Making
M. Makar, F. Johansson, J. Guttag, D. Sontag
ICML, 2020 [paper]
Recent News
Sept, 2022 | Our work on multi-shortcut removal, efficient offline RL, and concept credible learning got accpted to NeurIPS |
June, 2022 | Our papers on efficient RL, fairness and causality and causally motivated invariance have been accepted at ICML workshops! |
Jan, 2022 | Our paper on causally motivated shortcut removal was accepted at AISTATS! |
Sept, 2021 | I will be starting my position as a visitng researcher at MSR |
August, 2021 | I will be starting my position as the PPFP at UM |
July, 2021 | I will be presenting our work on predicting infectious diseases at ICML |
|
|