22-10-2012, 05:02 PM
Modeling No-Show Passengers on PACOM Exercise Airlift
ABSTRACT
The issue of no-shows has plagued many different industries that rely on reservation or appointment-type processes.
Everything from the restaurant to healthcare to commercial airline industry has had to come up with solutions to this problem
in order to remain both profitable and customer-service oriented. Many of these industries have used statistical models to
predict no-shows such that seats or appointments that would otherwise go unused are filled with customers that were
overbooked for those vacancies. This is a balancing act between trying to predict the number of no-shows and planning the
number to overbook, without having more customers show up than able to accommodate, an expensive proposition. The
military has been dealing with this same problem on its commercially chartered flights that move troops overseas for exercises,
contingencies, and other requirements. This research is aimed at using a statistical model to predict the number of no-shows
on chartered passenger airlift for Pacific Command joint exercises to try to maximize the utilization of seats left unused by
no-shows through overbooking techniques. If the number of no-shows can be predicted accurately, airlift planners will be able
to overbook missions to fill those seats and minimize wasted resources. This will be done using a purely quantitative approach
with correlation analysis and building a multiple regression model. The model will be built using both personnel and mission
characteristic factors from historical airlift data from major joint exercises in the Pacific Command’s area of responsibility.