SEMINAR: Decision Making Under Uncertainty: When Classical Assumptions Break Down
Guest: Yusuf Hakan Kalaycı, University of Southern California
Title: Decision Making Under Uncertainty: When Classical Assumptions Break Down (IE, CS, DS)
Date/Time: April 8, 2026, 10:40
Location: https://sabanciuniv.zoom.us/j/
Abstract: Decision-making under uncertainty underlies problems across economics and AI: pricing items to arriving buyers, selecting the best job candidate, or deciding when to stop sampling LLM responses. Classical theory offers elegant solutions but under idealized assumptions that rarely hold: that distributions are mutually independent, and known in advance. In this talk, I systematically study what happens when these assumptions break down, and show that strong algorithmic guarantees can survive — sometimes intact, sometimes at a provably unavoidable price. The scenarios we consider arise naturally in applications ranging from inference-time LLM optimization to optimal pricing in today's interconnected markets.
Bio: Yusuf Hakan Kalayci is a Ph.D. candidate in Computer Science at the University of Southern California, advised by Shaddin Dughmi and David Kempe. His research focuses on algorithm design for decision-making under uncertainty and computational social choice, with contributions published at STOC, ITCS, ICALP, AAAI, and AAMAS. He holds a B.Sc. in Computer Engineering and Mathematics with High Honors from Boğaziçi University, and is an IOI gold medalist.