Analyzing multi-factor effects on travel well-being, including non-linear relationship and interaction

Enhancing the travel well-being of commuters is crucial for the sustainable development of urban transportation and requires a clear understanding of the factors. However, existing research on the factors affecting travel well-being has not considered travel disturbance. This research adds travel disturbance and effort to a survey in Ningbo, China. Using this dataset, machine learning algorithms were employed to explore the complex relationship of seven variables on commuters' travel well-being. The results demonstrated machine learning algorithms such as Gradient Boosting Decision Tree and Random Forest outperform traditional linear regressions in analyzing travel well-being. The study identified built environment (Relative Importance = 24.6%) and affective effort (Relative Importance = 17.2%) were key determinants of travel well-being. Non-linear relationship between key variables and travel well-being was also investigated, and revealing a complex interaction between these variables. This research could help transportation managers provide more targeted and efficient suggestions to increase urban commuters' travel well-being.