Research map / Video / Video segmentation and tracking
Incomplete Label Multi-Task Deep Learning for Spatio-Temporal Event Subtype Forecasting
Yuyang Gao, Liang Zhao, Lingfei Wu, Yanfang Ye, Hui Xiong, Chaowei Yang
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 |
|---|---|---|---|
| add spatial regularization term (SIMDA-SR vs SR) (ablation) | Softmax regression · Brazil civil unrest | macro F1: 0.2247 -> 0.2560 | Improved |
| add spatial regularization term (SIMDA-SR vs SR) (ablation) | Softmax regression · China air pollution, 1-day | macro F1: 0.4287 -> 0.5572 | Improved |
| add spatial regularization (SIMDA-3 vs SBM-3) (ablation) | Shared-bottom model, 3 hidden layers · China air pollution, 1-day | macro F1: 0.5526 -> 0.6002 | Improved |
| add spatial regularization (SIMDA-1 vs SBM-1) (ablation) | Shared-bottom model, 1 hidden layer · Brazil civil unrest | macro F1: 0.2696 -> 0.2788 | Improved |
| add spatial regularization (SIMDA-1 vs SBM-1) (ablation) | Shared-bottom model, 1 hidden layer · China air pollution, 1-day | macro F1: 0.5812 -> 0.5560 | Hurt |
| add spatial regularization (SIMDA-3 vs SBM-3) (ablation) | Shared-bottom model, 3 hidden layers · Colombia civil unrest | macro F1: 0.2863 -> 0.2710 | Hurt |
| add spatial regularization (SIMDA-3 vs SBM-3) (ablation) | Shared-bottom model, 3 hidden layers · Venezuela civil unrest | macro F1: 0.2162 -> 0.2174 | Improved |
| add hidden layers (deep vs shallow models) | SIMDA/MLP/SBM · civil unrest and China air datasets | macro F1 | Improved |
| increase lead time from 1 to 7 days | SIMDA-3 · China air pollution | macro F1: 0.6002 -> 0.5425 | Hurt |
| use SIMDA over SBM | Shared-bottom model · Brazil civil unrest | AUC: 0.77 -> 0.81 | Improved |
| use hidden layers instead of softmax regression in SIMDA (ablation) | SIMDA-SR · input dimensionality scaling | runtime per ADMM step: around 26 seconds -> 6 seconds at 1300 dimensions | Improved |
Closest papers by meaning
- GeoPro-Net: Learning Interpretable Spatiotemporal Prediction Models Through Statistically-Guided Geo-Prototyping (AAAI 2025 · similarity 0.85)
- STD-PLM: Understanding Both Spatial and Temporal Properties of Spatial-Temporal Data with PLM (AAAI 2025 · similarity 0.85)
- FSTLLM: Spatio-Temporal LLM for Few Shot Time Series Forecasting (ICML 2025 · similarity 0.84)
- ST-FiT: Inductive Spatial-Temporal Forecasting with Limited Training Data (AAAI 2025 · similarity 0.83)
- Terra: A Multimodal Spatio-Temporal Dataset Spanning the Earth (NeurIPS 2024 · similarity 0.83)
- Graph-based Forecasting with Missing Data through Spatiotemporal Downsampling (ICML 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.
