Network theory has provided key insights into the overall resilience of air transportation systems. We expand upon these insights by using Mahalanobis distance to quantify delay abnormalities, complex network metrics for high-level insights, and a hybrid method that combines data-driven and network approaches. We apply these methods to public data and discuss trends in resilience among four US airlines. We find that our data-driven methods enable more detailed insights into airline resilience than traditional network methods. We also find that simultaneously considering all three approaches provides a more comprehensive understanding of resilience than the consideration of any one in isolation.

网络理论为航空运输系统的整体弹性提供了重要见识。我们使用Mahalanobis距离来量化延迟异常，用于高级洞察的复杂网络指标以及结合了数据驱动和网络方法的混合方法，来扩展这些洞察。我们将这些方法应用于公开数据，并讨论了美国四家航空公司的弹性趋势。我们发现，与传统的网络方法相比，我们的数据驱动方法可以更深入地了解航空公司的弹性。我们还发现，与单独考虑任何三种方法相比，同时考虑所有三种方法可提供对弹性的更全面的了解。