An Approximate Dynamic Programming Approach to Network Revenue Management with Customer Choice

A network revenue management problem is considered by the authors where customers use a prespecified choice model to choose among open fare products. The value function is approximated by the authors with an affine function of the state vector, starting with a Markov decision process (MDP) formulation. The authors show that a tighter bound for the MDP value is provided by the resulting problem than the choice-based linear program. A column generation algorithm is developed by to authors to solve the problem for a multinomial logit choice model with disjoint consideration sets. A bound is also derived by the authors as a decomposition heuristic byproduct. That the solution approach policies can signficantly outperform choice-based linear program heuristics is shown through this numerical study.

  • Availability:
  • Authors:
    • Zhang, Dan
    • Adelman, Daniel
  • Publication Date: 2009-8


  • English

Media Info

  • Media Type: Print
  • Features: Figures; References; Tables;
  • Pagination: pp 381-394
  • Serial:

Subject/Index Terms

Filing Info

  • Accession Number: 01140846
  • Record Type: Publication
  • Files: TRIS
  • Created Date: Sep 11 2009 12:21PM