Analysis of Dynamic Financial Performance of Airline Companies Using an Algorithm Combining Genetic Algorithm and DTW Clustering
Analysis of Dynamic Financial Performance of Airline Companies Using an Algorithm Combining Genetic Algorithm and DTW Clustering

Summary/Abstract: Air transport underpins modern mobility, trade and tourism making it a strategic pillar of the transport sector. Yet airlines often operate with fragile financial structures, underscoring the need for rigorous assessment of their financial performance. The literature features numerous studies that examine airlines’ financial performances and conduct mutual comparisons by constructing single composite indices based on financial indicators. To the best of the knowledge, no previous study has examined similarity and outlier detection among airlines based on the temporal dynamics of their financial performance indicators, nor has it conducted a financial analysis of the passenger airlines in Serbia. To accomplish the first objective, using financial statement data including 27 indicators for 15 airlines operating in the country for the 2021–2024 period, the study introduces purpose-built algorithm for the time-series-clustering. The algorithm integrates DTW-based k-means clustering and a Genetic Algorithm, identifies the most discriminative indicators between airline groups and yields sharper inter-cluster separation, and thereby more reliable similarity and outlier analysis than classical time-series clustering methods. Addressing the second objective, the evidence indicates recovery from the COVID-19 crisis in Serbia’s passenger air transport sector without sustained growth, calling for ongoing monitoring and targeted liquidity and resource allocation interventions.

• Details
• Contents