Optimization Online


Choice Based Revenue Management for Parallel Flights

Jim Dai(jim.dai***at***cornell.edu)
Weijun Ding(wding34***at***gatech.edu)
Anton Kleywegt(anton***at***isye.gatech.edu)
Xinchang Wang(xwang336***at***gatech.edu)
Yi Zhang(yzhang33***at***gatech.edu)

Abstract: This paper describes a revenue management project with a major airline that operates in a fiercely competitive market involving two major hubs and having more than 30 parallel daily flights. The market has a number of unusual characteristics including (1) almost half of customers choose not to purchase the tickets after booking; (2) about half of customers purchase their tickets within 3 days of departure; and (3) a significant number of customers no-show or go-show. We formulate choice based stochastic optimization problems to maximize the expected revenue for the airline. The inputs of the stochastic models include booking arrival rates, competitor assortment selection, booking choice probabilities for the airline's own flights as well as competitors' flights, booking-to-ticketing conversion probabilities, and go-show and no-show probabilities. We build a number of booking choice models, including multinomial logit models, nested logit models, and mixed logit models. The latter two types of models are aimed at incorporating unobserved heterogeneous customer preferences for different departure times. We formulate corresponding deterministic (fluid) optimization problems under each of the three booking choice models. We designed a column generation algorithm to compute optimal or near-optimal solutions for the deterministic problems, and the solutions are used to make assortment selections for the stochastic problem. The models used as input for the optimization problems are calibrated using 2011 data or 2012 data. Simulation studies using 2012 data show that the fluid based booking policies generate significantly more revenue than the airline's existing policy, and that policies based on the simpler multinomial logit models perform better than policies based on nested logit and mixed logit models, even when the simulation is based on the latter models, and even though the latter models seem to be more realistic.

Keywords: Pricing and Revenue Management, Transportation, Consumer Behavior

Category 1: Applications -- OR and Management Sciences

Category 2: Applications -- OR and Management Sciences (Yield Management )

Category 3: Applications -- OR and Management Sciences (Airline Optimization )

Citation: Technical report, School of Industrial and Systems Engineering, Georgia Institute of Technology, Atlanta, GA 30332-0205, USA

Download: [PDF]

Entry Submitted: 03/06/2014
Entry Accepted: 03/06/2014
Entry Last Modified: 03/06/2014

Modify/Update this entry

  Visitors Authors More about us Links
  Subscribe, Unsubscribe
Digest Archive
Search, Browse the Repository


Coordinator's Board
Classification Scheme
Give us feedback
Optimization Journals, Sites, Societies
Mathematical Optimization Society