Research map / Graph learning / GNN architectures
Kriging Convolutional Networks
Gabriel Appleby, Linfeng Liu, Li-Ping Liu
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 |
|---|---|---|---|
| feed neighbor training labels into GCN via KCN graph | GCN · Bird count | MSE: 0.70 ± .02 -> 0.50 ± .01 | Improved |
| feed neighbor labels (KCN-sage vs GraphSAGE) | GraphSAGE · Bird count | MSE: 0.53 ± .01 -> 0.44 ± .01 | Improved |
| replace kriging with KCN | Kriging · Bird count | MSE: 1.56 ± .85 -> 0.50 ± .01 | Improved |
| add dot-product attention to KCN (KCN-att) (ablation) | KCN · Bird count | MSE: 0.50 ± .01 -> 0.49 ± .01 | Improved |
| add dot-product attention to KCN (KCN-att) (ablation) | KCN · Yelp restaurant ratings | MSE: 0.990 ± .004 -> 0.977 ± .004 | No clear effect |
| replace GCN with KCN | GCN · Yelp restaurant ratings | MSE: 1.37 ± .006 -> 0.990 ± .004 | Improved |
| replace GraphSAGE with KCN-sage | GraphSAGE · Yelp restaurant ratings | MSE: 0.969 ± .004 -> 0.959 ± .004 | Improved |
| replace GCN with KCN | GCN · Precipitation | MSE: .640 ± .023 -> .029 ± .002 | Improved |
| replace kriging with KCN on log precipitation | Kriging · Precipitation | MSE: .155 ± .013 -> .029 ± .002 | Improved |
| replace GraphSAGE with KCN-sage | GraphSAGE · Precipitation | MSE: .056 ± .003 -> .030 ± .002 | Improved |
| use zero neighbors in KCN (no neighbor labels) (ablation) | KCN · Bird count | MSE | Hurt |
| use few neighbors in graph construction (ablation) | GCN · Bird count | MSE | Hurt |
Closest papers by meaning
- Inductive Graph Neural Networks for Spatiotemporal Kriging (AAAI 2021 · similarity 0.91)
- KITS: Inductive Spatio-Temporal Kriging with Increment Training Strategy (AAAI 2025 · similarity 0.89)
- Graph-Time Convolutional Neural Networks: Architecture and Theoretical Analysis (TPAMI 2023 · similarity 0.88)
- FastGCN: Fast Learning with Graph Convolutional Networks via Importance Sampling (ICLR 2018 · similarity 0.85)
- pathGCN: Learning General Graph Spatial Operators from Paths (ICML 2022 · similarity 0.85)
- Generalizing CNNs to graphs with learnable neighborhood quantization (NeurIPS 2024 · similarity 0.84)
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.
