Research map / Label-efficient and robust learning / Recognition and long-tail
Spatiality Preservable Factored Poisson Regression for Large-Scale Fine-Grained GPS-Based Population Analysis
Masamichi Shimosaka, Yuta Hayakawa, Kota Tsubouchi
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 graph-proximity spatiality preservation regularizer (ablation) | SPF · Tokyo GPS 100 m mesh | MNLL: 1.47 -> 1.46 | Improved |
| add graph-proximity spatiality preservation regularizer (ablation) | SPF · Tokyo GPS 200 m mesh | MNLL: 2.86 -> 2.83 | Improved |
| add graph-proximity spatiality preservation regularizer (ablation) | SPF · Tokyo GPS 1 km mesh | MNLL: 21.94 -> 21.27 | Improved |
| factorize weights across meshes instead of independent per-mesh models | bilinear Poisson regression · Tokyo GPS 100 m mesh | MNLL: 1.60 -> 1.46 | Improved |
| factorize weights across meshes instead of independent per-mesh models | bilinear Poisson regression · Tokyo GPS 600 m mesh | MNLL: 9.91 -> 9.89 | No clear effect |
| use per-mesh patterns instead of one shared model for all meshes | bilinear Poisson regression · Tokyo GPS 1 km mesh | MNLL: 23.7 -> 21.27 | Improved |
| use shared pattern factorization (SPF) instead of low-rank mesh-index model | BP with index · Tokyo GPS 100 m mesh | MNLL: 1.48 -> 1.46 | Improved |
| use factorized parameters instead of per-mesh models | bilinear Poisson regression · Tokyo GPS | training time | Improved |
Closest papers by meaning
- SeoulMMOD: A Large-Scale Multimodal Origin-Destination Flow Benchmark (NeurIPS 2026 · similarity 0.82)
- Taming the Long Tail in Human Mobility Prediction (NeurIPS 2024 · similarity 0.82)
- Learning from Highly Sparse Spatio-temporal Data (NeurIPS 2024 · similarity 0.82)
- GeoRanker: Distance-Aware Ranking for Worldwide Image Geolocalization (NeurIPS 2025 · similarity 0.82)
- Satellites Reveal Mobility: A Commuting Origin-destination Flow Generator for Global Cities (NeurIPS 2025 · similarity 0.82)
- RDPI: A Refine Diffusion Probability Generation Method for Spatiotemporal Data Imputation (AAAI 2025 · similarity 0.81)
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.
