Fanny Yang
Fanny Yang is an associate professor at ETH Zurich whose research studies when machine-learning models generalize and how they can remain reliable under data shifts. She earned degrees at Karlsruhe Institute of Technology and Technical University Munich, completed a PhD at UC Berkeley, and joined ETH's computer science department after roles at Amazon AWS, ETH's Institute for Theoretical Studies, and Stanford. Her work connects statistical learning theory with early stopping for boosting and privacy-fairness trade-offs in machine learning. She received tenure at ETH in 2026.
- Now at
- ETH Zurich
- Badges
Timeline
-
2007–2013
Degrees in Germany
She earns an undergraduate degree at Karlsruhe Institute of Technology in 2010 and an M.Sc. at Technical University Munich in 2013.
-
2013–2018
A Berkeley PhD
She completes a PhD in UC Berkeley's EECS department in 2018 under Martin Wainwright. Her research there explores computational and statistical aspects of machine-learning methods for generalization.
-
2017
Early stopping for boosting
She coauthors a 2017 NeurIPS Spotlight paper with Yuting Wei and Martin J. Wainwright on early stopping for kernel boosting. It links stopped estimates to localized Gaussian complexity and derives optimal stopping rules for kernel classes.
-
June 2017–February 2018
Applied science at AWS
She works as an Applied Scientist Intern at Amazon AWS in Palo Alto from June 2017 to February 2018.
-
September 2018–December 2019
Postdoctoral roles at ETH and Stanford
She is a Junior Fellow (postdoc) at ETH's Institute for Theoretical Studies, working with Nicolai Meinshausen, from September 2018 to September 2019. At Stanford, she works with John Duchi and Percy Liang as a postdoctoral fellow in Statistics and Computer Science from April 2019 to December 2019.
-
January 2020
Joining ETH Zurich
She joins ETH Zurich's Department of Computer Science as a tenure-track assistant professor in January 2020. Her research examines the theoretical performance and robust generalization of machine-learning models, and when their predictions can be trusted.
-
2022
Privacy and fairness at UAI
With Amartya Sanyal and Yaxi Hu, she coauthors 'How unfair is private learning?', which UAI lists in a 2022 oral session. For long-tailed data, their paper shows that private learning can constrain higher accuracy on minority subpopulations. It also shows that relaxing overall accuracy can improve fairness under strict privacy.
-
As of October 2026
Associate professor at ETH
She is an associate professor in ETH Zurich's Department of Computer Science and receives tenure in 2026. Her research focuses on statistical learning and reliable, robust machine learning.
Career
- Since 2026ETH ZurichAssociate Professor
- 2019Stanford UniversityPostdoctoral Fellow in Statistics and Computer Science
- 2017–2018AmazonApplied Scientist Intern
Education
- 2018University of California, BerkeleyPhD
- 2013Technical University of MunichM.Sc.
- 2010Karlsruhe Institute of TechnologyUndergraduate degree
Colleagues
Sources and credits
Photo: We have no published photo of Fanny Yang yet. Send us a free-licensed photo.