TY - JOUR

T1 - Using Machine Learning Techniques to Assess the Financial Impact of the COVID-19 Pandemic on the Global Aviation Industry

AU - Halteh, Khaled

AU - AlKhoury, Ritab

AU - Adel Ziadat, Salem

AU - Gepp, Adrian

AU - Kumar, Kuldeep

N1 - Open Access to once it is published it will be available.

PY - 2024/3

Y1 - 2024/3

N2 - Prediction of financial distress is a crucial concern for decision-makers, especially in industries prone to external shocks, such as the aviation sector. This study employs machine learning techniques on a comprehensive global dataset of aviation companies to develop highly accurate financial distress prediction models. These models empower stakeholders with informed decision-making capabilities to navigate the aviation industry's challenges, most notably exemplified by the COVID-19 pandemic. The aviation industry holds substantial economic importance, contributing significantly to revenue, employment, and economic activity worldwide. However, its susceptibility to external factors underscores the need for robust predictive tools. Leveraging advances in machine learning, this study pioneers the application of data-driven, non-parametric solutions to the aviation sector, both before and after the pandemic. Importantly, this study addresses a gap in the field by conducting comparative evaluations of prediction models, which have been lacking in previous research efforts, often leading to inconclusive outcomes. Key findings of the study highlight the Random Forest and Stochastic Gradient Boosting models as the most accurate in forecasting financial distress within the aviation industry. Notably, the study identifies debt-to-equity, return on invested capital, and debt ratio as the most important predictors of financial distress in this context.

AB - Prediction of financial distress is a crucial concern for decision-makers, especially in industries prone to external shocks, such as the aviation sector. This study employs machine learning techniques on a comprehensive global dataset of aviation companies to develop highly accurate financial distress prediction models. These models empower stakeholders with informed decision-making capabilities to navigate the aviation industry's challenges, most notably exemplified by the COVID-19 pandemic. The aviation industry holds substantial economic importance, contributing significantly to revenue, employment, and economic activity worldwide. However, its susceptibility to external factors underscores the need for robust predictive tools. Leveraging advances in machine learning, this study pioneers the application of data-driven, non-parametric solutions to the aviation sector, both before and after the pandemic. Importantly, this study addresses a gap in the field by conducting comparative evaluations of prediction models, which have been lacking in previous research efforts, often leading to inconclusive outcomes. Key findings of the study highlight the Random Forest and Stochastic Gradient Boosting models as the most accurate in forecasting financial distress within the aviation industry. Notably, the study identifies debt-to-equity, return on invested capital, and debt ratio as the most important predictors of financial distress in this context.

KW - Aviation industry

KW - Financial distress prediction

KW - Machine Learning

KW - COVID-19

U2 - 10.1016/j.trip.2024.101043

DO - 10.1016/j.trip.2024.101043

M3 - Article

JO - Transportation Research Interdisciplinary Perspectives

JF - Transportation Research Interdisciplinary Perspectives

SN - 2590-1982

M1 - 101043

ER -