Research map / Label-efficient and robust learning / Recognition and long-tail
Hyperbolic Procrustes Analysis Using Riemannian Geometry
Ya-Wei Eileen Lin, Yuval Kluger, Ronen Talmon
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 |
|---|---|---|---|
| apply HPA alignment instead of no alignment | Lorentz-model embedding · BC batch removal | k-NN AUC-ROC: 0.5254 ± 0.0091 -> 0.7410 ± 0.0354 | Improved |
| apply HPA alignment instead of no alignment | Lorentz-model embedding · LC batch removal | k-NN AUC-ROC: 0.5521 ± 0.0991 -> 0.8316 ± 0.0904 | Improved |
| apply HPA alignment instead of no alignment | Lorentz-model embedding · CyTOF P1 BT | k-NN AUC-ROC: 0.6646 ± 0.1556 -> 0.9401 ± 0.0068 | Improved |
| use HPA instead of Riemannian translation RT alone (ablation) | RT · BC batch removal | k-NN AUC-ROC: 0.6001 ± 0.0658 -> 0.7410 ± 0.0354 | Improved |
| use HPA instead of Riemannian translation RT alone (ablation) | RT · LC batch removal | k-NN AUC-ROC: 0.5318 ± 0.0370 -> 0.8316 ± 0.0904 | Improved |
| use HPA instead of RT alone (ablation) | RT · simulated alignment | pairwise discrepancy | No clear effect |
| use HPA instead of optimal transport HOT-ME | HOT-ME · LC batch removal | k-NN AUC-ROC: 0.4901 ± 0.0516 -> 0.8316 ± 0.0904 | Improved |
| use HPA instead of HOT-ME | HOT-ME · BC batch removal | MMD: 0.0007 ± 0.0002 -> 0.0013 ± 0.0004 | Hurt |
| use HPA instead of HOT-F | HOT-F · CyTOF P1 BT | MMD: 0.0004 ± 0.0002 -> 0.0012 ± 0.0002 | Hurt |
| apply HPA instead of no alignment | Lorentz-model embedding · BC batch removal | MMD: 0.2089 ± 0.0027 -> 0.0013 ± 0.0004 | Improved |
| use HPA instead of HOT-L under noise and distortion | HOT-L · simulated alignment | pairwise discrepancy | Improved |
| compare aligned multi-batch HPA to single-batch classification | HPA · LC batch removal | k-NN AUC-ROC: 0.9137 ± 0.0532 -> 0.8316 ± 0.0904 | Hurt |
Closest papers by meaning
- HyperTools: a Python Toolbox for Gaining Geometric Insights into High-Dimensional Data (JMLR 2018 · similarity 0.85)
- CO-SNE: Dimensionality Reduction and Visualization for Hyperbolic Data (CVPR 2022 · similarity 0.85)
- Fast Hyperboloid Decision Tree Algorithms (ICLR 2024 · similarity 0.83)
- Hyperbolic Dataset Distillation (NeurIPS 2025 · similarity 0.82)
- Probabilistic Geometric Principal Component Analysis with application to neural data (ICLR 2025 · similarity 0.82)
- Geometry-Aware Similarity Metrics for Neural Representations on Riemannian and Statistical Manifolds (NeurIPS 2026 · similarity 0.82)
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.
