Quantifying delay propagation in airline networks
We develop a framework for quantifying delay propagation in airline networks by integrating structural modeling and machine learning methods to estimate causal effects. Using a comprehensive dataset on actual delays and a model-selection algorithm (elastic net), we estimate a weighted directed graph...
Saved in:
| Main Authors: | , , |
|---|---|
| Format: | text |
| Language: | English |
| Published: |
Institutional Knowledge at Singapore Management University
2024
|
| Subjects: | |
| Online Access: | https://ink.library.smu.edu.sg/soe_research/2795 https://ink.library.smu.edu.sg/context/soe_research/article/3794/viewcontent/Quantifying_Delay_Propagation_in_Airline_Networks__1_.pdf |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Institution: | Singapore Management University |
| Language: | English |
| Summary: | We develop a framework for quantifying delay propagation in airline networks by integrating structural modeling and machine learning methods to estimate causal effects. Using a comprehensive dataset on actual delays and a model-selection algorithm (elastic net), we estimate a weighted directed graph of delay propagation for each major airline in the United States and establish conditions under which the propagation coefficients are causal. These estimates enable a decomposition of airline performance into "luck" and "ability." Our findings indicate that luck accounts for approximately 38% of the performance difference between Delta and American Airlines in our data. Additionally, we leverage these estimates to analyze how network topology and other airline characteristics, such as aircraft fleet heterogeneity, influence expected delays. |
|---|