Research map / Learning theory and algorithms / Clustering
A Simple Approach to Automated Spectral Clustering
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 relative-eigen-gap instead of eigen-gap as selection criterion (ablation) | AutoSC-GD · Yale B | clustering accuracy: 0.790 -> 0.897 | Helped |
| use relative-eigen-gap instead of eigen-gap (ablation) | AutoSC-GD · Fashion-MNIST | clustering accuracy: 0.562 -> 0.595 | Helped |
| use relative-eigen-gap instead of J(L,-10) regularizer (ablation) | AutoSC-GD · AR | clustering accuracy: 0.832 -> 0.786 | Hurt |
| use kernel LSR with polynomial kernel instead of epsilon-neighborhood AMC within AutoSC-BO (ablation) | AutoSC-BO · Yale B first 10 subjects | clustering accuracy: 0.325 -> 0.966 | Helped |
| add automated selection of AMC and hyperparameters via relative eigen-gap (AutoSC-GD) over best-tuned baselines | spectral clustering · Yale B | clustering accuracy: 0.806 -> 0.897 | Helped |
| remove NSE (neural eigenspace extension) from AutoSC-GD (ablation) | AutoSC-GD+NSE · MNIST features 10k | clustering accuracy: 0.9772 vs lower with NSE; time cost increases | Mixed |
| use ReLU instead of tanh in NSE hidden layer (ablation) | AutoSC+NSE · MNIST | clustering accuracy | Helped |
| increase number of landmark points s (ablation) | AutoSC+NSE · MNIST | clustering accuracy | Helped |
Closest papers by meaning
- One-Step Spectral Clustering via Dynamically Learning Affinity Matrix and Subspace (AAAI 2017 · similarity 0.91)
- A Tighter Analysis of Spectral Clustering, and Beyond (ICML 2022 · similarity 0.90)
- Scalable Sequential Spectral Clustering (AAAI 2016 · similarity 0.90)
- Equivalence between Graph Spectral Clustering and Column Subset Selection (Student Abstract) (AAAI 2024 · similarity 0.89)
- Pass-Efficient Algorithms for Graph Spectral Clustering (Student Abstract) (AAAI 2024 · similarity 0.88)
- Structure-Aware Spectral Sparsification via Uniform Edge Sampling (NeurIPS 2025 · similarity 0.87)
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.