A Primal-Dual Approach Toward Resource-Constrained Revenue Management with Demand Learning and Large Action Space

This paper proposes an approach that can be applied to solve several important revenue management (RM) problems with demand learning and potentially large action space constrained by initial unreplenishable resources. This approach combines the technique of the primal-dual method in optimization and upper confidence bound algorithm in learning. Three important RM problems are studied in this paper: network revenue management, dynamic assortment selection with a multinomial-logit choice model, and joint pricing and assortment optimization problems. This paper demonstrates how application-specific subroutines can be developed for each revenue management problem, achieving state-of-the-art theoretical performance guarantees with computationally efficient algorithms. Numerical experiments are conducted to demonstrate that our algorithms have great empirical performance and they are computationally efficient.

Funding: S. Miao gratefully acknowledges financial support provided by the Ruegg Family Faculty Scholar Award and the Leeds School of Business.

Supplemental Material: All supplemental materials, including the code, data, and files required to reproduce the results, are available at https://doi.org/10.1287/opre.2021.0483.