Transportation

A measurement model of pedestrian tolerance time under signal-controlled conditions

X. Hu, N. Wang, et al.

Discover how a novel stacking model developed by Xinghua Hu and colleagues predicts pedestrian red light tolerance times! By utilizing advanced machine learning techniques, this research outshines traditional models and offers valuable insights into optimizing red light durations for different pedestrian groups.

00:00

Playback language: English

Index

Abstract

This paper proposes a stacking model for predicting pedestrian red light tolerance time. The model uses XGBoost, random forest, support vector machine regression, and multilayer perceptron as primary models, with multiple linear regression as a meta-model. Data was categorized into low, medium, and high tolerance groups using random survival forest and K-means clustering. The grouped stacking model outperformed individual models, with varying MSE, MAE, and MAPE across tolerance groups. The method helps determine optimal red light durations at crossings based on pedestrian composition.

Publisher

HUMANITIES AND SOCIAL SCIENCES COMMUNICATIONS

Published On

Aug 31, 2024

Authors

Xinghua Hu, Nanhao Wang, Jiahao Zhao, Xiaochuan Zhou, Bing Long

Related Publications

Explore these studies to deepen your understanding of the subject.

Economics

Analysing the heterogeneity in working conditions of migrant informal workers in China: a test of the WIEGO model of informal employment

G. Huang, B. Cai, et al.

Medicine and Health

Safety and reactogenicity of a controlled human infection model of sand fly-transmitted cutaneous leishmaniasis

V. Parkash, H. Ashwin, et al.

Health and Fitness

Effect of time restricted eating on body weight and fasting glucose in participants with obesity: results of a randomized, controlled, virtual clinical trial

P. M. Peeke, F. L. Greenway, et al.

Medicine and Health

Pulse transit time-estimated blood pressure: a comparison of beat-to-beat and intermittent measurement

S. Hoshide, A. Yoshihisa, et al.

Listen, Learn & Level Up

Over 10,000 hours of research content in 25+ fields, available in 12+ languages.

No more digging through PDFs, just hit play and absorb the world's latest research in your language, on your time.

listen to research audio papers with researchbunny