Periodic-CRN: A Convolutional Recurrent Model for Crowd Density Prediction with Recurring Periodic Patterns
Abstract
Time-series forecasting in geo-spatial domains has important applications, including urban planning, traffic management and behavioral analysis. We observed recurring periodic patterns in some spatio-temporal data, which were not considered explicitly by previous non-linear works. To address this lack, we propose novel `Periodic-CRN' (PCRN) method, which adapts convolutional recurrent network (CRN) to accurately capture spatial and temporal correlations, learns and incorporates explicit periodic representations, and can be optimized with multi-step ahead prediction. We show that PCRN consistently outperforms the state-of-the-art methods for crowd density prediction across two taxi datasets from Beijing and Singapore.
Cite
Text
Zonoozi et al. "Periodic-CRN: A Convolutional Recurrent Model for Crowd Density Prediction with Recurring Periodic Patterns." International Joint Conference on Artificial Intelligence, 2018. doi:10.24963/IJCAI.2018/519Markdown
[Zonoozi et al. "Periodic-CRN: A Convolutional Recurrent Model for Crowd Density Prediction with Recurring Periodic Patterns." International Joint Conference on Artificial Intelligence, 2018.](https://mlanthology.org/ijcai/2018/zonoozi2018ijcai-periodic/) doi:10.24963/IJCAI.2018/519BibTeX
@inproceedings{zonoozi2018ijcai-periodic,
title = {{Periodic-CRN: A Convolutional Recurrent Model for Crowd Density Prediction with Recurring Periodic Patterns}},
author = {Zonoozi, Ali and Kim, Jung-Jae and Li, Xiao-Li and Cong, Gao},
booktitle = {International Joint Conference on Artificial Intelligence},
year = {2018},
pages = {3732-3738},
doi = {10.24963/IJCAI.2018/519},
url = {https://mlanthology.org/ijcai/2018/zonoozi2018ijcai-periodic/}
}