Estimation of High-Dimensional Contextual Pricing Models with Nonparametric Price Confounders

• 成果类型: Article; Early Access
• 署名作者: Wang, Yining; Liu, Quanquan
• 署名单位: University of Texas System; University of Texas Dallas
• 刊物名称: OPERATIONS RESEARCH
• ISSN/ISSBN: 0030-364X
• DOI: 10.1287/opre.2024.1002
• 发表日期: 2025
• 关键词: selection; demand; Lasso

摘要

Personalized pricing with contextual information is a widespread practice in a number of revenue management problems. A pricing algorithm or platform utilizes users' personal data to make the most profitable pricing decisions, which could vary among individuals. In this paper, we study the question of estimating a contextual demand regression model with high-dimensional data, incorporating an unknown, nonparametric pricing function that acts as a confounding term to the demand model. We propose a high-dimensional instrumental variable regression method that uses properly centered contextual vectors as approximate instruments in a Lasso (Least absolute shrinkage and selection operator) formulation to mitigate the bias from price confounders. We further propose a debiased approach based on two-level partitioning of the price interval from dynamic programming and show that the debiased approach typically results in smaller estimation errors. Finally, we corroborate our methodological and theoretical results with numerical studies and propose some questions for future research.