Contextual Pricing: Adversarial Corruptions

This lecture explores contextual search in the presence of irrational agents, focusing on online markets and sequential decision-making. It delves into challenges like unbounded computation, personalization needs, and unknown priors, emphasizing the impact of adversarial corruptions in behavioral models. The instructor presents a novel algorithmic framework for contextual pricing, addressing both known and unknown corruption levels. The lecture concludes by discussing the broader implications of adversarial corruptions in contextual pricing and online advertising, highlighting the need for robust algorithms in modern data-driven systems.