Using machine learning to analyze and predict entry patterns of low-cost airlines: A study of Southwest Airlines

Abstract

This paper uses machine learning to analyze and predict entry patterns of Southwest Airlines into various city pair. The purpose is to understand the parameters impacting the decision to enter into a city pair, by a low cost airline. Decision to enter (exit) from a market depends on endogenous factors and exogenous factors, such as decisions of other airlines, passenger profile, airport profile, competition on that sector, global economic conditions. The paper uses supervised machine learning to understand and predict a low cost airlines decision to enter (exit) a specific city pair. The uniqueness of this paper is that this kind of prediction was done for the first time using the existing data pre-processing and machine learning techniques. Moreover, an analysis of Southwest’s entry patterns would help (1) competing airlines to better plan their networks, (2) non-competing airlines to learn about how to plan their entry patterns and (3) airports to better plan their operations and slots. The analysis clearly showed a shift in Southwest’s entry and exit strategy over the last several years.