Scholarlink

This paper presents a comparative analysis of Multinomial Logit (MNL) and Mixed Multinomial Logit (MMNL) models for evaluating mode choice behavior at the University of Zagreb Borongaj Campus. The MMNL model, by incorporating random parameters for travel time and cost, reveals significant heterogeneity in user preferences that the simpler MNL model fails to capture. Results show that while travel time, cost, and vehicle crowding negatively influence mode choice in both models, the MMNL model provides a superior fit and richer behavioral interpretation. Its enhanced performance is evidenced by improved log-likelihood and goodness of fit measures. These findings emphasize the importance of modeling individual-level variation in transport studies. Future research may benefit from further preference segmentation using latent class models or socio-demographic interactions to better inform targeted transport policies.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Modeling Mode Choice to University Campus Using Discrete Choice Methods

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Nikola Kožul,
  • Luka Novačko,
  • Marjana Petrović,
  • Karlo Babojelić

摘要

This paper presents a comparative analysis of Multinomial Logit (MNL) and Mixed Multinomial Logit (MMNL) models for evaluating mode choice behavior at the University of Zagreb Borongaj Campus. The MMNL model, by incorporating random parameters for travel time and cost, reveals significant heterogeneity in user preferences that the simpler MNL model fails to capture. Results show that while travel time, cost, and vehicle crowding negatively influence mode choice in both models, the MMNL model provides a superior fit and richer behavioral interpretation. Its enhanced performance is evidenced by improved log-likelihood and goodness of fit measures. These findings emphasize the importance of modeling individual-level variation in transport studies. Future research may benefit from further preference segmentation using latent class models or socio-demographic interactions to better inform targeted transport policies.