Scholarlink

<p>The public bicycle system (PBS) has become an essential complement to transit services and a popular choice for short trips. Despite this, there is limited research on the nonlinear relationships between the built environment and PBS usage, as well as the interaction effects among various built environment factors. This study investigates PBS usage in Taipei using the Extreme Gradient Boosting (XGBoost) model and SHapley Additive exPlanations (SHAP). The analysis reveals complex nonlinear relationships and threshold effects between built environment characteristics and PBS usage. Notably, land use mix emerges as a key factor, while distance to the city center and proximity to public transportation also significantly impact PBS usage. Additionally, the study reveals intricate interaction effects among built environment features. These findings offer valuable insights for designing urban environments that better support cycling.</p>

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

Nonlinear and interaction effects of the built environment on public bicycle usage: a case study of Taipei City非线性及建筑环境对公共自行车使用的影响：台北市的案例研究

  • 期刊论文

  • 发表日期：

    2026年7月8日

查看全文
  • Yajuan Zhang,
  • Xiaoyan Huang

摘要

The public bicycle system (PBS) has become an essential complement to transit services and a popular choice for short trips. Despite this, there is limited research on the nonlinear relationships between the built environment and PBS usage, as well as the interaction effects among various built environment factors. This study investigates PBS usage in Taipei using the Extreme Gradient Boosting (XGBoost) model and SHapley Additive exPlanations (SHAP). The analysis reveals complex nonlinear relationships and threshold effects between built environment characteristics and PBS usage. Notably, land use mix emerges as a key factor, while distance to the city center and proximity to public transportation also significantly impact PBS usage. Additionally, the study reveals intricate interaction effects among built environment features. These findings offer valuable insights for designing urban environments that better support cycling.