A systematic comparative evaluation of machine learning classifiers and discrete choice models for travel mode choice in the presence of response heterogeneity
作者：
Highlights：
• Systematic comparison between logit models and five machine learning methods.
• Discrete choice models for travel mode with response heterogeneity are used.
• Prediction accuracy and variable importance for each travel mode is investigated.
• XGB, RF and NN outperform traditional logit models in predicting travel mode choices.
• The feature importance method recognized the non-informative variables for each mode.
摘要
•Systematic comparison between logit models and five machine learning methods.•Discrete choice models for travel mode with response heterogeneity are used.•Prediction accuracy and variable importance for each travel mode is investigated.•XGB, RF and NN outperform traditional logit models in predicting travel mode choices.•The feature importance method recognized the non-informative variables for each mode.
论文关键词：Travel choice model,Machine learning,Classification,Discrete choice models,Feature importance,Multinomial logit model,Mixed multinomial logit model
论文评审过程：Received 27 July 2020, Revised 25 July 2021, Accepted 15 November 2021, Available online 1 January 2022, Version of Record 11 January 2022.
论文官网地址：https://doi.org/10.1016/j.eswa.2021.116253
