Hyperparameter Optimisation and Comparative Analysis of Machine Learning Models for Travel Mode Choice Prediction

Understanding the determinants of travel mode choice (TMC) in urban contexts is essential for effective transport planning and policy development. Past studies predominantly employed traditional discrete choice models because of their simplicity, diversity, and high interpretability; however, they rely on restrictive assumptions. Although machine learning (ML) techniques have shown promising predictive capabilities, comparative assessments of traditional and ML approaches, particularly considering hyperparameter optimisation, remain limited. This study addresses this gap by comparing a traditional model with four ML algorithms: decision tree (DT), random forest (RF), support vector machine (SVM), and k-nearest neighbour (KNN). In addition, systematic hyperparameter optimisation is performed to evaluate its impact on predictive performance relative to default model settings. Feature importance analysis is also conducted to identify the key determinants of TMC. The analysis is based on a multi-dimensional, three-week household time-use and activity diary dataset comprising 508 individuals from 191 households in the Bandung Metropolitan Area, Indonesia. The results demonstrate that ML models outperform traditional methods, while hyperparameter optimisation substantially improves model performance across all considered algorithms against default models. Notably, the KNN model exhibits a 16.67% increase in accuracy, followed by the SVM model with an 11.15% improvement. Among the optimised evaluated models, SVM achieves the best overall performance, with a macro-averaged accuracy of 0.588 and a precision of 0.591. Feature importance analysis reveals that total travel time is the most influential determinant of TMC. These findings highlight the importance of model tuning and hyperparameter optimisation in ML-based TMC prediction and provide insights into the factors shaping travel behaviour. The outcomes can support more informed decision-making in urban transport planning and policy formulation.