A prediction interval framework-based spatial–temporal convolution block network for traffic demand prediction

doi:10.1016/j.tre.2025.104426

A prediction interval framework-based spatial–temporal convolution block network for traffic demand prediction

2025-09-20

0

PRE

AI

DOI：10.1016/j.tre.2025.104426

Original

Original request for help

Save

Abstract

En 中文

• Propose a novel travel demand interval prediction model based on deep learning and prediction interval (PI) framework. • Develop spatiotemporal feature extraction mechanism that integrates irregular region division and a pre-training process. • Design an ingenious loss function for PI framework. • Validate the superiority of the proposed method compared to advanced models in a real dataset. • Interpret the results of the proposed method to potential traffic applications.

A prediction interval framework-based spatial–temporal convolution block network for traffic demand prediction

Abstract

Journal

Organization