Improved Forecast Accuracy in Airline Revenue Management by Unconstraining Demand Estimates from Censored Data

Front Cover
Universal-Publishers, 2001 - Business & Economics - 276 pages
Accurate forecasts are crucial to a revenue management system. Poor estimates of demand lead to inadequate inventory controls and sub-optimal revenue performance. Forecasting for airline revenue management systems is inherently difficult. Competitive actions, seasonal factors, the economic environment, and constant fare changes are a few of the hurdles that must be overcome. In addition, the fact that most of the historical demand data is censored further complicates the problem. This dissertation examines the challenge of forecasting for an airline revenue management system in the presence of censored demand data. This dissertation analyzed the improvement in forecast accuracy that results from estimating demand by unconstraining the censored data. Little research has been done on unconstraining censored data for revenue management systems. Airlines tend to either ignore the problem or use very simple ad hoc methods to deal with it. A literature review explores the current methods for unconstraining censored data. Also, practices borrowed from areas outside of revenue management are adapted to this application. For example, the Expectation-Maximization (EM) and other imputation methods were investigated. These methods are evaluated and tested using simulation and actual airline data. An extension to the EM algorithm that results in a 41% improvement in forecast accuracy is presented.
 

What people are saying - Write a review

We haven't found any reviews in the usual places.

Contents

2810 The EM Algorithm for Unconstraining Censored Demand Data
80
2811 Details of the EStep for the EM Algorithm
84
2812 Details of the MStep for the EM Algorithm
86
2813 Numerical Illustration for the EM Algorithm
88
2814 Projection Detruncation Method
91
2815 Details of the EStep for the PD Method
96
2816 Details of the MStep for the PD Method
98
2817 Numerical Illustration for the PD Algorithm
99

171 FirstCome FirstServed
18
172 Leglevel
19
173 Virtual Nesting
20
174 Origin and Destination Itinerary Level
22
18 Optimization Methods
23
181 Expected Marginal Seat Revenue EMSR
24
182 Network Formulations
27
183 Deterministic Linear Program
28
184 Probabilistic Nonlinear Program
29
Forecasting Methods Literature Review and Current Practices
30
21 Types of Forecasting
32
22 Macrolevel Forecasting
33
23 Passenger Choice Modeling
34
241 Exponential Smoothing
38
242 Moving Average
39
243 Linear Regression
40
244 Additive Pickup Model
41
245 Multiplicative Pickup Model
46
25 Censored Data
48
26 Cost of Using Censored Data
49
261 Sensitivity Analysis
51
27 Methods for Handling Incomplete Data
52
271 CompleteData Methods
53
272 Imputation Methods
54
273 Statistical Model Methods
55
28 Unconstraining Censored Data
56
281 The Goal of Unconstraining Censored Data
57
282 Capture the Data Directly
62
283 Ignore the Censored Data
63
284 Discard the Censored Data
64
285 Mean Imputation Method
66
286 Median Imputation Method
69
287 Percentile Imputation Method
71
288 Multiplicative Booking Profile Method
73
289 ExpectationMaximization EM Algorithm
78
Modeling the Censored Data Problem
102
31 Censored Data Simulation
104
311 Demand Generation and Data Collection
105
312 Simulate Censoring of the Data
107
32 Simulation Validation
116
33 Design of Experiments
119
332 Experimental Units
123
333 Performance Measurements
124
334 Definition of Constrained Data
128
335 Number of Replications
131
34 Experiment Procedure
133
35 Distribution Analysis
136
Analysis and Comparison of Unconstraining Methods
145
41 Experiment Results for the Ignore Method
147
42 Experiment Results for the Discard Method
156
43 Experiment Results for the Mean Imputation Method
167
44 Experiment Results for the Median Imputation Method
177
45 Experiment Results for the Percentile Imputation Method
186
46 Experiment Results for the Booking Profile Method
194
47 Experiment Results for the ExpectationMaximization Algorithm
206
471 Rate of Convergence for the EM Algorithm
215
48 Extended EM Algorithm
217
481 Extension to the EM Algorithm
219
482 Experiment Results for the Extended EM Algorithm
221
49 Experiment Results for the ProjectionDetruncation Algorithm
230
491 Sensitivity to Tau
239
410 Comparison of Unconstraining Methods Performance
241
4101 A Note on the Performance Results
245
Conclusion
247
51 Contributions
248
52 Future Research Directions
249
522 Overestimates of Demand
251
53 Implementation Issues
252
Bibliography
254
Copyright

Common terms and phrases

Popular passages

Page 9 - Airlines (1987) defined the goal of yield management as "to maximize passenger revenue by selling the right seats to the right customers at the right time." 4 Pfiefer (1989) described airline yield management as the "process by which discount fares are allocated to scheduled flights for the purposes of balancing demand and increasing revenues.

Bibliographic information