Research map / Graph learning / Graph algorithms and clustering
Latent Smooth Skeleton Embedding
Li Wang, Qi Mao, Ivor Tsang
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 |
|---|---|---|---|
| PSL-l1 versus Kmeans on raw data | PSL-ℓ1 · COIL20 clustering | accuracy: 0.6674 -> 0.7319 | Improved |
| PSL-l1 versus Kmeans on raw data | PSL-ℓ1 · USPS clustering | accuracy: 0.6153 -> 0.7309 | Improved |
| PSL-l1 versus Kmeans on raw data | PSL-ℓ1 · YALE-B clustering | accuracy: 0.1081 -> 0.3521 | Improved |
| PSL-l2 versus Kmeans on raw data | PSL-ℓ2 · Isolet clustering | accuracy: 0.5633 -> 0.6794 | Improved |
| Learn a sparse graph (PSL-l1) instead of fixed neighborhood graph (PSL-l2) | PSL-ℓ1 · Isolet clustering | accuracy: 0.6794 -> 0.5973 | Hurt |
| PSL-l1 with learned sparse graph versus PSL-l2 with fixed neighborhood (ablation) | PSL-ℓ1 · YALE-B clustering | NMI: 0.6094 -> 0.4796 | Hurt |
| PSL-l1 versus PSL-l2 (ablation) | PSL-ℓ1 · Letter clustering | accuracy: 0.2816 -> 0.3320 | Improved |
| PSL-l1 versus SMCE (sparse graph learning baseline) | PSL-ℓ1 · Pendigits clustering | NMI: 0.8071 -> 0.8230 | Improved |
| PSL-l2 versus SMCE | PSL-ℓ2 · COIL20 clustering | accuracy: 0.3813 -> 0.7181 | Improved |
| Probabilistic posterior modeling (PSL-l1) versus GPLVM | PSL-ℓ1 · USPS clustering | accuracy: 0.4709 -> 0.7309 | Improved |
| PSL-l1 versus MEU | PSL-ℓ1 · Satimage clustering | accuracy: 0.7398 -> 0.7454 | Improved |
| PSL versus landmark MVU (lMVU) | PSL-ℓ1 · COIL20 clustering | accuracy: 0.5042 -> 0.7319 | Improved |
Closest papers by meaning
- Probabilistic Structural Latent Representation for Unsupervised Embedding (CVPR 2020 · similarity 0.86)
- Probabilistic Dimensionality Reduction via Structure Learning (TPAMI 2017 · similarity 0.85)
- Euclidean Embedding of Data Using Local Distances (NeurIPS 2026 · similarity 0.85)
- Reliable and Efficient Anytime Skeleton Learning (AAAI 2020 · similarity 0.85)
- Denoising Functional Maps: Diffusion Models for Shape Correspondence (CVPR 2025 · similarity 0.84)
- Partition First, Embed Later: Laplacian-Based Feature Partitioning for Refined Embedding and Visualization of High-Dimensional Data (ICML 2025 · 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.
