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

被引:129
|
作者
Zhang, Dan [1 ]
Adelman, Daniel [2 ]
机构
[1] McGill Univ, Desautels Fac Management, Montreal, PQ H3A 1G5, Canada
[2] Univ Chicago, Grad Sch Business, Chicago, IL 60637 USA
关键词
network revenue management; choice behavior; dynamic programming; MODEL; ALLOCATION; DIVERSION; FLIGHTS;
D O I
10.1287/trsc.1090.0262
中图分类号
C93 [管理学]; O22 [运筹学];
学科分类号
070105 ; 12 ; 1201 ; 1202 ; 120202 ;
摘要
We consider a network revenue management problem where customers choose among open fare products according to some prespecified choice model. Starting with a Markov decision process (MDP) formulation, we approximate the value function with an affine function of the state vector. We show that the resulting problem provides a tighter bound for the MDP value than the choice-based linear program. We develop a column generation algorithm to solve the problem for a multinomial logit choice model with disjoint consideration sets (MNLD). We also derive a bound as a by-product of a decomposition heuristic. Our numerical study shows the policies from our solution approach can significantly outperform heuristics from the choice-based linear program.
引用
收藏
页码:381 / 394
页数:14
相关论文
共 50 条