Optimization with Costly Information (in person only)

Modern algorithms increasingly rely on data to make decisions, but acquiring data can itself be costly. In this talk, we consider a class of selection problems with stochastic inputs, where an algorithm may learn about each input random variable only through a sequence of costly actions. The goal is to determine what information to acquire before optimizing the downstream objective. This framework generalizes the classical Pandora’s Box problem.

I will describe new results that connect costly-information optimization to well-studied models in stochastic online selection, including prophet inequalities and the query-commit model. This talk is based on joint work with Dimitris Christou and Trung Dang.