I am a researcher at Microsoft Research Cambridge working on deep generative models for inverse problems and analog-amenable machine learning. I did my PhD in machine learning with Bernhard Schölkopf at the Max Planck Institute for Intelligent Systems. Previously, I studied mathematics and physics in Cambridge and Frankfurt.
During my PhD I developed training algorithms for robust and causal machine learning based on empirical likelihood estimation. I also worked on inverse problems and temporal point processes.
You can find me on Github, LinkedIn and Google Scholar.
News
- (Juli 2024) I’m at LPHYS’24 in Brazil to present our work on Analog Optical Computing for Machine Learning and Optimization—the next-gen hardware accelerator developed by our team at MSR Cambridge
- (May 2024) Our paper Geometry-Aware Instrumental Variable Regression got accepted at ICML 2024
- (March 2024) Joined Microsoft Research as a machine learning researcher
- (Feb 2024) Submitted my PhD Thesis on Empirical Likelihood Estimators for Robust and Causal Learning
- (May 2023) Our paper Estimation Beyond Data Reweighting: Kernel Method of Moments got accepted at ICML 2023
- (May 2023) Gave a talk on Distributionally Robust Machine Learning with Kernel Machines at SIAM Optimization 23 in Seattle
- (Feb 2023) Our paper Compact holographic sound fields enable rapid one-step assembly of matter in 3D got published at Science Advances
- (Dec 2022) Gave an invited talk on Distributionally Robust Machine Learning via Conditional Moment Restrictions at the Weierstrass Institute for Applied Analysis and Stochastics, Berlin
- (July 2022) Our paper Maximum Mean Discrepancy Distributionally Robust Nonlinear Chance-Constrained Optimization with Finite-Sample Guarantee got accepted at CDC 2022
- (May 2022) Our paper Functional Generalized Empirical Likelihood Estimation for Conditional Moment Restrictions got accepted at ICML 2022