Air Travel Itinerary Market Share Prediction using Artificial Intelligence

Abstract

The widely used Quality Service Index (QSI) model estimates air traffic by considering attributes such as capacity, connectivity, and travel time, which impact an itinerary market share (MS) in the origin-destination (OD) market. To determine itinerary attractiveness, the conventional QSI model combines these features, weighting them based on their significance and generates a score for each option. The key is therefore to select the optimal combination of weighted attributes that are able to forecast the MS with the desired accuracy. However, traditional QSI models rely on expert knowledge and assume that the traffic generation and allocation is always based in the same market principles, hindering their ability to generalize and identify new features. To address this, the present work, which is part of a research project entitled ERA funded by Red.es, introduces a generalized rationale based on Artificial intelligence using historical data, a challenge that has been faced in the early stages of the project. This approach enables the identification and understanding of key features in the aviation traffic and the development of a forecasting model capable of capturing more complex feature relationships.