Research map / Learning theory and algorithms / Clustering
Bregman Power k-Means for Clustering Exponential Family Data
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 Bregman divergence with power-mean annealing instead of Euclidean power k-means | Power k-Means · Binomial simulation | ARI: 0.915 (0.004) -> 0.931 (0.003) | Helped |
| Use Bregman divergence with power-mean annealing instead of Euclidean power k-means | Power k-Means · Poisson simulation | ARI: 0.888 (0.006) -> 0.916 (0.004) | Helped |
| Use Bregman divergence with power-mean annealing instead of Euclidean power k-means | Power k-Means · Gamma simulation | ARI: 0.677 (0.008) -> 0.879 (0.004) | Helped |
| Use power-mean annealing with Bregman divergence instead of hard Bregman clustering | Bregman Hard Clustering · Binomial simulation | ARI: 0.886 (0.011) -> 0.931 (0.003) | Helped |
| Use power-mean annealing with Bregman divergence instead of hard Bregman clustering | Bregman Hard Clustering · Poisson simulation | ARI: 0.882 (0.010) -> 0.916 (0.004) | Helped |
| Use power-mean annealing with Bregman divergence instead of hard Bregman clustering | Bregman Hard Clustering · Gamma simulation | ARI: 0.868 (0.005) -> 0.879 (0.004) | Helped |
| Use Bregman power means rather than Euclidean power means | Power k-Means · Gaussian simulation | ARI: 0.927 (0.003) -> 0.927 (0.003) | No clear effect |
| Set initial power to -3 instead of -0.2 (ablation) | Bregman Power k-Means · Poisson simulation, p=50 | ARI: 0.442 (0.002) -> 0.924 (0.003) | Helped |
| Set initial power to -9 instead of -3 (ablation) | Bregman Power k-Means · Poisson simulation, p=2 | ARI: 0.412 (0.005) -> 0.321 (0.006) | Hurt |
Closest papers by meaning
- Clustering with Bregman Divergences: an Asymptotic Analysis (NeurIPS 2016 · similarity 0.92)
- Modified K-means Algorithm with Local Optimality Guarantees (ICML 2025 · similarity 0.88)
- Power k-Means Clustering (ICML 2019 · similarity 0.86)
- Learning to Approximate a Bregman Divergence (NeurIPS 2020 · similarity 0.86)
- Clustering Semi-Random Mixtures of Gaussians (ICML 2018 · similarity 0.85)
- Approximation Bounds for Hierarchical Clustering: Average Linkage, Bisecting K-means, and Local Search (NeurIPS 2017 · similarity 0.85)
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.