Risk Minimization as a Framework for Online Allocation in Display Advertising

• 成果类型: Article
• 署名作者: Shamsi, Davood; Luenberger, Robert; Ye, Yinyu
• 署名单位: Shanghai Jiao Tong University; Shanghai Institute for Mathematics & Interdisciplinary Sciences
• 刊物名称: OPERATIONS RESEARCH
• ISSN/ISSBN: 0030-364X
• DOI: 10.1287/opre.2024.0737
• 发表日期: 2026
• 关键词: online algorithms; linear programming; primal-dual; dynamic price update; risk minimization

摘要

This research note revisits the framework proposed in our earlier work and explores its conceptual and algorithmic connection to recent advances in dual-based online resource allocation-particularly the dual mirror descent method introduced previously. Both approaches address the challenge of making real-time sequential allocation decisions under dynamically revealed constraints. Although the dual mirror descent method relies on Bregman divergence to guide dual updates, our framework derives nearly identical exponential update rules through a convex risk minimization lens. We show that this alternative perspective not only recovers known allocation strategies-including greedy and linear-but also offers a flexible, interpretable foundation for designing robust online algorithms. By highlighting the mathematical parallels and modeling distinctions between these paradigms, we aim to broaden the theoretical toolkit for online optimization and motivate further study of risk-aware dual methods.