Research map / Label-efficient and robust learning / Zero-/few-shot and meta-learning
Task-Adaptive Meta-Learning Framework for Advancing Spatial Generalizability
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 an easy-to-hard task hierarchy with models adapted to task difficulty | MAML for county-level corn yield prediction · Corn yield, whole yield counties | R² (%): 79.01 -> 80.58 | Helped |
| Use an easy-to-hard task hierarchy with hard-task training | MAML for county-level corn yield prediction · Corn yield, low-yield counties | R² (%): 60.62 -> 64.39 | Helped |
| Train Easy and Hard models with 3 inner epochs instead of 1 (ablation) | Adaptive-MAML · Corn yield, whole yield counties | R² (%): 80.58 -> 79.77 | Hurt |
| Train Easy and Hard models with 3 inner epochs instead of 1 (ablation) | Adaptive-MAML · Corn yield, low-yield counties | R² (%): 64.39 -> 63.81 | Hurt |
| Increase test-time adaptation from 1 to 2 steps (ablation) | Adaptive-MAML-B · Corn yield, low-yield counties | R² (%): 64.39 -> 68.52 | Helped |
| Increase test-time adaptation from 1 to 2 steps (ablation) | Adaptive-MAML-B · Corn yield, whole yield counties | R² (%): 80.58 -> 80.57 | No clear effect |
| Increase test-time adaptation from 1 to 2 steps (ablation) | Adaptive-MAML-A · Corn yield, whole yield counties | R² (%): 79.77 -> 77.98 | Hurt |
| Increase the number of inner training epochs (ablation) | Adaptive-MAML · Corn yield prediction | R² | Hurt |
| Set the maximum split number to 1, 2, or 3 (ablation) | Adaptive-MAML-B · Corn yield prediction | R² | No clear effect |
Closest papers by meaning
- Get Rid of Isolation: A Continuous Multi-task Spatio-Temporal Learning Framework (NeurIPS 2024 · similarity 0.87)
- Referee-Meta-Learning for Fast Adaptation of Locational Fairness (AAAI 2024 · similarity 0.85)
- Diffusion Transformers as Open-World Spatiotemporal Foundation Models (NeurIPS 2025 · similarity 0.84)
- Unifying Knowledge from Diverse Datasets to Enhance Spatial-Temporal Modeling: A Granularity-Adaptive Geographical Embedding Approach (ICML 2025 · similarity 0.83)
- Task Groupings Regularization: Data-Free Meta-Learning with Heterogeneous Pre-trained Models (ICML 2024 · similarity 0.83)
- Drawing Informative Gradients from Sources: A One-stage Transfer Learning Framework for Cross-city Spatiotemporal Forecasting (AAAI 2025 · 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.