Optimal Pricing with a Single Point
Video
Team Information
Team Members
Achraf Bahamou, PhD Candidate, Industrial Engineering and Operations Research, School of Engineering and Applied Science, Columbia University
Amine Allouah, Research Scientist, Decision, Risk & Operations, Graduate School of Business, Columbia University
Faculty Advisor: Omar Besbes, Decision, Risk & Operations, Graduate School of Business, Columbia University
Abstract
In the present work, we study the following fundamental data-driven pricing problem: How can/should a decision-maker price its product based on observations at a single historical price? The decision-maker optimizes over randomized pricing policies the worst-case ratio of the revenue it can garner compared to an oracle with full knowledge of the distributions of values, when the distribution are only assumed to belong to broad non-parametric set. In particular, our framework applies to the widely used regular and monotone hazard rate classes of distributions. For settings where the seller knows the exact percentile associated to a price or only a confidence interval for it, we fully characterize optimal performance and near-optimal (randomized) pricing algorithms that adjust to the information at hand. The framework we develop is general and allows to characterize optimal deterministic mechanisms, incorporate uncertainty in the percentile, and additional data at other prices.
Contact this Team
Team Contact: Achraf Bahamou (use form to send email)