Behavioral Analysis of Urban Travel Mode Selection Based on Random Forest Algorithm

Assessing the travel demand involves a knowledge of people make choices with regard to mode of transportation. It was stated that machine learning (ML) techniques are beneficial for predicting achievement and additionally suggested for modeling mode choice patterns. Nevertheless, establishing an effective rationale for the association between inputs and outputs is challenging because of the black-box structure of ML. This research examines the predictability and interpretability of the mathematical framework by analyzing travel mode selections using an innovative adaptive travel modeoptimized random forest (AWPO-RF) approach. The prediction performance of the RF is enhanced by implementing the AWPO strategy. The predictive efficacy of the suggested technique is investigated using Python and trip journal information which has been gathered. The experimental outcomes show that the suggested strategy outperformed other existing strategies for more accurate travel mode selection estimate. Moreover, the AWPO-RF technique that has been proposed to computes the relative importance of explanatory variables and their association with mode selections. This was necessary for understanding and realistic modelling of travelling behaviors.