Web sitemizin güncellenme işlemleri devam etmektedir.

Personalized Dynamic Pricing with Machine Learning

Bora Keskin
Assoc. Prof.

Bora Keskin is an Associate Professor in the Operations Management area at the Fuqua School of Business at Duke University. Bora received his B.S. in Industrial Engineering and Mathematics from Bogazici University in 2007, and his Ph.D. from the Graduate School of Business at Stanford University in 2012. Before joining the faculty at Duke University in 2015, he worked at McKinsey & Company as a consultant in banking and telecommunications industries, and at the University of Chicago as an Assistant Professor of Operations Management.

Bora's main research studies management problems that involve decision making under uncertainty. In particular he is interested in stochastic models and their application to revenue management, dynamic pricing, statistical learning, machine learning, and product differentiation. Bora has published papers in leading research journals such as Management Science, Operations Research, Manufacturing and Service Operations Management, and Mathematics of Operations Research. He serves as an associate editor and as a referee for these journals and has received several Distinguished and Meritorious Service Awards. In 2019, Bora was awarded the Lanchester Prize for the development of a novel paradigm for the modeling and analysis of online dynamic optimization problems that are subject to temporal uncertainty.

Bora has taught Value Chain Innovation in Business Processes as well as Supply Chain Management for the Daytime and Executive MBA programs, and Revenue Management for the PhD program at the Fuqua School of Business. Outside Duke, he served as a Board Member for the INFORMS Revenue Management and Pricing (RM&P) Section from 2014-2016, and as a Cluster Chair for the RM&P and M&SOM-Service tracks at INFORMS Annual Meetings (organizing 319 talks in total).

Abstract

We consider a seller who can dynamically adjust the price of a product at the individual customer level, by utilizing information about customers' characteristics encoded as a d-dimensional feature vector. We assume a personalized demand model, parameters of which depend on s out of the d features. The seller initially does not know the relationship between the customer features and the product demand, but learns this through sales observations over a selling horizon of T periods. We prove that the seller’s expected regret, i.e., the revenue loss against a clairvoyant who knows the underlying demand relationship, is at least of order sT^{1/2} under any admissible policy. We then design a near-optimal pricing policy for a “semi-clairvoyant” seller (who knows which s of the d features are in the demand model) that achieves an expected regret of order sT^{1/2}log(T). We extend this policy to a more realistic setting where the seller does not know the true demand predictors, and show this policy has an expected regret of order sT^{1/2}[log(d)+log(T)], which is also near-optimal. Finally, we test our theory on simulated data and on a data set from an online auto loan company in the United States. On both data sets, our experimentation-based pricing policy is superior to intuitive and/or widely-practiced customized pricing methods such as myopic pricing and segment-then-optimize policies. Furthermore, our policy improves upon the loan company’s historical pricing decisions by 47% in expected revenue over a six-month period.

Bora Keskin
Kasım 23, 2020 - 18:00