Data-Intensive Distributed Systems LaboratoryCS Seminar
Date: February 18th, 2026
Time: 12:50pm - 1:45pm
Room: SB111
Dr. Yutong Wang
Assistant Professor
Computer Science Department
Illinois Institute of Technology
Talk Title
Towards Privacy-Preserving Universal Approximators via Tropical Rational Functions
Talk Abstract
This work develops the theory and practice for learning with tropical rational functions, which are directly implementable in a tropical fully homomorphic encryption scheme. Fully homomorphic encryption (FHE) schemes allow clients to use machine learning models on the cloud without compromising privacy for privacy-sensitive applications in areas such as healthcare. State-of-the-art FHE schemes that are applied to deep learning are approximate and introduce numerical imprecision. Moreover, handling non-linearities such as ReLU remains a challenge for FHE schemes. Towards overcoming both of these two drawbacks, we develop the universal approximation theory and memory-efficient training layer for learning with tropical rational functions.
Speaker Bio
Yutong Wang is an Assistant Professor in the Computer Science Department at Illinois Institute of Technology, where he leads the EMC2 (Elements of MultiClass Classification) Lab. He is also a member of the IDEAL Institute. His research focuses on the theoretical and practical aspects of classification, including uncertainty estimation, calibration, and privacy-preserving machine learning. He is the recipient of an NSF CRII award on uncertainty estimation and robustness in hierarchical classification. Previously, Yutong was a Postdoctoral Research Fellow supported by the Eric and Wendy Schmidt AI Fellowship in at the University of Michigan, Ann Arbor. He completed his Ph.D. in Electrical and Computer Engineering at the University of Michigan and obtained his Masters in Mathematics at the University of California, Davis.