Research map / Learning theory and algorithms / Clustering
Low-Rank Doubly Stochastic Matrix Decomposition for Cluster Analysis
Zhirong Yang, Jukka Corander, Erkki Oja
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 |
|---|---|---|---|
| Use DCD instead of k-means | clustering · MNIST | purity: 0.48 -> 0.97 | Improved |
| Use DCD instead of best baseline on large data | clustering · 19 data sets with N > 4500 | best-method count: 17 and 11 cases in purity and NMI | Improved |
| Use DCD instead of 1-Spec | clustering · MNIST | purity: 0.88 -> 0.97 | Improved |
| Use DCD instead of LSC | clustering · 20NG | purity: 0.50 -> 0.62 | Improved |
| Use DCD instead of Ncut | clustering · TDT2 | purity: 0.89 -> 0.86 | Hurt |
| Use DCD instead of k-means | clustering · CITESEER | purity: 0.68 -> 0.50 | Hurt |
| Use DCD instead of LSD | clustering · NORB | purity: 0.42 -> 0.35 | Hurt |
| Select cluster number by NOSAC residual instead of CH-km | cluster number selection · MNIST | selected number of clusters (truth 10): 2 -> 10 | Improved |
| Select cluster number by NOSAC residual instead of gap-km | cluster number selection · COIL100 | selected number of clusters (truth 100): 140 -> 79 | Improved |
| Use 5NN instead of 10NN input similarity (ablation) | DCD · COIL20 | selected number of clusters (truth 20): 17 -> 21 | Improved |
| Use 5NN instead of 10NN input similarity (ablation) | DCD · COIL100 | selected number of clusters (truth 100): 79 -> 99 | Improved |
Closest papers by meaning
- Graph-based Clustering Revisited: A Relaxation of Kernel k-Means Perspective (JMLR 2026 · similarity 0.89)
- Statistically Optimal $K$-means Clustering via Nonnegative Low-rank Semidefinite Programming (ICLR 2024 · similarity 0.89)
- Learning-Based Low-Rank Approximations (NeurIPS 2019 · similarity 0.88)
- Robust Bi-stochastic Graph Regularized Matrix Factorization for Data Clustering (TPAMI 2020 · similarity 0.88)
- Clustering Ensemble Meets Low-rank Tensor Approximation (AAAI 2021 · similarity 0.88)
- Low-Rank Riemannian Optimization for Graph-Based Clustering Applications (TPAMI 2021 · similarity 0.88)
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.
