Utilizing a machine learning ensemble to evaluate the service quality and passenger satisfaction among public transportations

Public transportation is an essential criterion that benefits several social sectors. Hence, most developing countries display an increase in the demand for enhanced public utility vehicle (PUV) systems. PUVs are prevalent in the Philippines; however, research on passenger satisfaction and public transportation is scarce. This research aimed to assess passengers' future intentions regarding PUVs through passenger satisfaction utilizing various latent variables. This study utilized an online survey with a total of 600 respondents that are using PUVs in the Philippines who voluntarily answered the questionnaire. The data were analyzed using different Machine Learning Algorithms (MLA) such as Deep Learning Neural Network (DLNN), Decision Tree (DT), and Random Forest Classifier (RFC). The study indicated that people vastly prefer a route-efficient way of traveling, safety, value for money, and passenger expectations as it highly affected passenger satisfaction and future intentions. The theoretical basis of this study provided an effective instrument for resolving the country's emerging traffic issues and served as the foundation for forming PUVs and policy initiatives. Future research may look into and concentrate more on particular types of service quality factors and public utility vehicle to provide a more in-depth analysis of the subject and extend the analysis. Researchers may also utilize MLA for the data as it provides a more efficient and accurate factor analysis in the transportation sector. Finally, managerial insights could be elevated, including service domains in different areas.