Research map / Learning theory and algorithms / Clustering
KNNDA: A New Perspective of Alignment Recovery for Partially View-Aligned Clustering
Liang Zhao, Tianqi Yue, Shubin Ma, Ziyue Wang, ZhiYuan Liu, Bo Xu
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 KNNDA alignment module instead of Hungarian-algorithm alignment (ablation) | KNNDA · Scene-15 | ACC (%): 27.25 -> 45.08 | Improved |
| add KNNDA alignment module instead of Hungarian-algorithm alignment (ablation) | KNNDA · HandWritten | ACC (%): 73.15 -> 91.20 | Improved |
| add alignment confidence score and threshold filtering (ablation) | KNNDA · Scene-15 | ACC (%): 31.22 -> 45.08 | Improved |
| add alignment confidence score and threshold filtering (ablation) | KNNDA · HandWritten | ACC (%): 82.05 -> 91.20 | Improved |
| restrict alignment recovery to K-nearest-neighbor alignment domains (ablation) | KNNDA · all datasets | memory | Improved |
| raise fixed threshold above 0.5 on BDGP (ablation) | KNNDA · BDGP | clustering metrics | Hurt |
| raise fixed threshold (ablation) | KNNDA · BDGP and ALOI-100 | Class Alignment Rate | Improved |
| use larger threshold decay step (ablation) | KNNDA dynamic threshold · BDGP and ALOI-100 | clustering performance | Hurt |
| vary number of neighbors K from 5 to 20 (ablation) | KNNDA · BDGP | clustering performance | No clear effect |
| add a third view | KNNDA · 100Leaves (3V vs 2V) | ACC (%): 70.38 -> 83.06 | Improved |
| use KNNDA over best baseline on Scene-15 NMI | KNNDA · Scene-15 | NMI (%): 43.93 -> 42.55 | Hurt |
Closest papers by meaning
- Learning Whom to Align With: Progressive Anomaly Combination Detection for Partially View-Aligned Clustering (AAAI 2026 · similarity 0.93)
- Scalable Cross-View Sample Alignment for Multi-View Clustering with View Structure Similarity (NeurIPS 2025 · similarity 0.92)
- AF-UMC: An Alignment-Free Fusion Framework for Unaligned Multi-View Clustering (NeurIPS 2025 · similarity 0.92)
- Robust Multi-view Clustering against Imperfect Information (NeurIPS 2026 · similarity 0.89)
- KOALA: Kernel Coupling and Element Imputation Induced Multi-View Clustering (AAAI 2025 · similarity 0.89)
- Collaborative Similarity Fusion and Consistency Recovery for Incomplete Multi-view Clustering (AAAI 2025 · similarity 0.89)
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.
