Name
Title | ||
|---|---|---|
Temporal Attention-Based Graph Convolution Network for Taxi Demand Prediction in Functional Areas |
Abstract | ||
|---|---|---|
Shared travel is increasingly becoming an indispensable way of urban transportation. Accurately calculating the demand for taxis in various regions of the city has become a huge problem. In this paper, we divide the city into multiple lattices of different sizes and propose a graph convolution network based on the temporal attention mechanism for taxi demand prediction in each functional area of the city. The model includes graph convolution network (GCN), temporal convolution network (TCN), and the attention mechanism, which are respectively used to capture the spatial correlation of roads, time dependence, and highlight the characteristics of the time-series data. Extensive experiments on three datasets validate the effectiveness of the proposed method, compared against several state-of-the-art methods. Despite there are amount differences among the three datasets in our experiment, our model still has a high prediction accuracy. Our model code is available at https://github.com/qdu318/TAGCN. |
Year | DOI | Venue |
|---|---|---|
2021 | 10.1007/978-3-030-85928-2_16 | WIRELESS ALGORITHMS, SYSTEMS, AND APPLICATIONS, WASA 2021, PT I |
Keywords | DocType | Volume |
Taxi demand prediction, Attention mechanism, Graph convolution network, Temporal convolution network | Conference | 12937 |
ISSN | Citations | PageRank |
0302-9743 | 0 | 0.34 |
References | Authors | |
0 | 5 | |
Name | Order | Citations | PageRank |
|---|---|---|---|
| Yue Wang | 1 | 260 | 40.39 |
| Jianbo Li | 2 | 46 | 28.87 |
| Aite Zhao | 3 | 16 | 4.37 |
| Zhiqiang Lv | 4 | 26 | 11.28 |
| Guangquan Lu | 5 | 0 | 0.34 |