Title: Data-driven Inverse Optimization with Imperfect Information
|Affilliation:||University of Waterloo|
We continue our reading group with a paper by Kuhn et al with the same title as above. In fact, in data-driven inverse optimization an observer aims to learn the preferences of an agent who solves a parametric optimization problem depending on an exogenous signal. Thus, the observer seeks the agent's objective function that best explains a historical sequence of signals and corresponding optimal actions. The paper focuses on situations where the observer has imperfect information, that is, where the agent's true objective function is not contained in the search space of candidate objectives, where the agent suffers from bounded rationality or implementation errors, or where the observed signal-response pairs are corrupted by measurement noise.