Join us in welcoming Zhan Pang, professor at Purdue, as he discusses optimistic online learning and dynamic pricing. Alums and friends are always welcome to attend all ISE seminars.

Join via Zoom
https://ncsu.zoom.us/j/93197460841?pwd=WALamWThFuPLR8czVn1Evo30DZf735.1
Meeting ID: 931 9746 0841
Passcode: 964529

Title and Abstract

Optimistic Online Learning and Dynamic Pricing

We introduce Optimistic Revenue Maximizing Pricing (ORMP), an optimism-driven online decision rule that incorporates statistical uncertainty directly into dynamic pricing. Adhering to the principle of Optimism in the Face of Uncertainty (OFU), ORMP embeds decision-level optimism by constructing an optimistic revenue envelope over the price space. This envelope is obtained by regularizing the sample-average revenue rate with a price-dependent exploration bonus derived from time-uniform confidence sequences, and prices are chosen to maximize this envelope in an online fashion.
This envelope-based formulation makes optimism explicit in pricing behavior and provides a structural interpretation for systematic markups, boundary pricing, and persistent price variations under uncertainty. The exploration bonus may convexify the optimistic revenue envelope, inducing more dispersed pricing experimentation and controlled exploration across the price space. The algorithm leverages time-uniform confidence sequences to achieve finite-time performance guarantees that remain valid under adaptive pricing policies and random stopping horizons. We further show the induced pricing paths and performance trajectories admit direct interpretation and post-hoc auditing from realized prices, establishing a transparent link between exploration-exploitation tradeoffs and observed revenue outcomes. Simulation studies demonstrate that ORMP consistently achieves significant revenue improvements relative to the standard benchmark of certainty-equivalent (greedy) pricing policy, underscoring the practical and theoretical appeal of optimistic decision rule for dynamic pricing and revenue management.

Biography

Zhan Pang is a faculty member in Supply Chain and Operations Management and a Purdue University Faculty Scholar. He is the coordinator of the doctoral program in Supply Chain and Operations Management (SCOM). His research interests include statistical learning and decision theory, healthcare delivery systems, supply chain risk management, and pricing and revenue management. He is a senior editor for Production and Operations Management , associate editor for IISE Transactions, and a founding editor of Journal of Blockchain Research. He has rich industry experience as entrepreneur and management consultant. Since he joined Purdue, he has formed research and experiential learning partnerships with many corporations including Dow Chemical, Corteva and Wabash. He is also working closely with healthcare service organizations including a non-for-profit home healthcare service provider to help them improve efficiency and quality of care. He has served the board of directors for China Titans Energy Technology Group Co., Ltd. from 2015-2023. As one of Purdue’s Innovation and Entrepreneurship Fellows, he works with other fellows across campus to help connect Purdue innovators with novel paths for entrepreneurship and commercialization to take the innovations to the world.