Personalized travel time estimation for urban road networks: A tensor-based context-aware approach

作者：

Highlights：

• A novel tensor-based approach for travel time modeling is proposed.

• A variety of contextual features are extracted and utilized.

• A context-aware and collaborative estimation objective function is developed.

• A gradient-based algorithm to find an optimal solution is devised.

• The proposed model is evaluated by a real case study in Beijing, China.

摘要

•A novel tensor-based approach for travel time modeling is proposed.•A variety of contextual features are extracted and utilized.•A context-aware and collaborative estimation objective function is developed.•A gradient-based algorithm to find an optimal solution is devised.•The proposed model is evaluated by a real case study in Beijing, China.

论文关键词：Data driven,Time slot,Dynamic,Block term decomposition,Sparse,GPS trajectory

论文评审过程：Received 13 November 2017, Revised 25 January 2018, Accepted 25 February 2018, Available online 9 March 2018, Version of Record 20 March 2018.

论文官网地址：https://doi.org/10.1016/j.eswa.2018.02.033