Research map / Graph learning / Knowledge and temporal graphs
Semi-Supervised Hierarchical Recurrent Graph Neural Network for City-Wide Parking Availability Prediction
Weijia Zhang, Hao Liu, Yanchi Liu, Jingbo Zhou, Hui Xiong
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| Use SHARE instead of DCRNN | City-wide parking availability prediction · Beijing and Shenzhen, 15/30/45 min | MAE and RMSE: Beijing MAE: (31.8%, 31.3%, 29.9%); RMSE: (30.9%, 31.5%, 30.9%). Shenzhen MAE: (29.6%, 28.7%, 30.5%); RMSE: (28.8%, 28.8 | Improved |
| Add PA approximation to CxtGNN (ablation) | CxtGNN, yielding CAGNN · Beijing, 15/30/45 min | MAE and RMSE: MAE: (18.6%, 18.6%, 18.3%) lower; RMSE: (12.2%, 12.6%, 13.5%) lower | Improved |
| Add soft clustering graph convolution to CAGNN (ablation) | CAGNN, yielding SHARE · Beijing, 15/30/45 min | MAE and RMSE: MAE: (14.2%, 14.1%, 13.4%) lower; RMSE: (20%, 19.7%, 18.2%) lower | Improved |
| Increase labeled parking-lot ratio from 0.1 to 0.9 (ablation) | SHARE · Beijing | MAE | Improved |
| Increase latent-node ratio from 0.01 to 0.1 (ablation) | SHARE · Beijing | MAE | Improved |
| Increase latent-node ratio from 0.1 to 0.8 (ablation) | SHARE · Beijing | MAE | Hurt |
| Set input history length to 12 time steps rather than other tested lengths (ablation) | SHARE · Beijing | MAE | Improved |
| Increase prediction horizon from 1 to 6 steps (ablation) | SHARE · Beijing | prediction error | Hurt |
Closest papers by meaning
- Adaptive Graph Convolutional Recurrent Network for Traffic Forecasting (NeurIPS 2020 · similarity 0.85)
- Spatial-Temporal Synchronous Graph Convolutional Networks: A New Framework for Spatial-Temporal Network Data Forecasting (AAAI 2020 · similarity 0.85)
- Spatiotemporal Multi-Graph Convolution Network for Ride-Hailing Demand Forecasting (AAAI 2019 · similarity 0.85)
- Learning Heterogeneous Spatial-Temporal Representation for Bike-Sharing Demand Prediction (AAAI 2019 · similarity 0.85)
- I Open at the Close: A Deep Reinforcement Learning Evaluation of Open Streets Initiatives (AAAI 2024 · similarity 0.84)
- Towards Dynamic Spatial-Temporal Graph Learning: A Decoupled Perspective (AAAI 2024 · similarity 0.83)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
