Multi-task and multi-view learning based on particle swarm optimization for short-term traffic forecasting

作者：

Highlights：

• Multi-task and multi-view feature learning models combined for traffic forecasting.

• Prediction results of three spatiotemporal views obtained using a set of kernels.

• Added regularization terms used to constrain all tasks to select shared features.

• Proposed model outperforms existing baseline methods in prediction accuracy.

摘要

•Multi-task and multi-view feature learning models combined for traffic forecasting.•Prediction results of three spatiotemporal views obtained using a set of kernels.•Added regularization terms used to constrain all tasks to select shared features.•Proposed model outperforms existing baseline methods in prediction accuracy.

论文关键词：Multi-view learning,Multi-task learning,Particle swarm optimization,Spatiotemporal dependency,Spatiotemporal heterogeneity,Task relationship learning

论文评审过程：Received 21 September 2018, Revised 2 May 2019, Accepted 15 May 2019, Available online 24 May 2019, Version of Record 12 June 2019.

论文官网地址：https://doi.org/10.1016/j.knosys.2019.05.023