Research map / Learning theory and algorithms / Clustering
Deep Divergence Learning
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 |
|---|---|---|---|
| Replace Euclidean distance with a learned Deep Bregman divergence under triplet loss | Deep metric learning · Cifar10 | k-nn classification accuracy: 77.00 -> 81.40 | Helped |
| Replace Euclidean distance with a learned Deep Bregman divergence under triplet loss | Deep metric learning · Fashion MNIST | k-nn classification accuracy: 93.24 -> 94.90 | Helped |
| Replace Euclidean distance with a learned Deep Bregman divergence under triplet loss | Deep metric learning · SVHN | k-nn classification accuracy: 92.58 -> 94.03 | Helped |
| Replace Euclidean distance with a learned Deep Bregman divergence under triplet loss | Deep metric learning · STL10 | k-nn classification accuracy: 59.97 -> 62.64 | Helped |
| Replace Euclidean distance with a learned Deep Bregman divergence under triplet loss | Deep metric learning · MNIST | k-nn classification accuracy: 99.50 -> 99.61 | Helped |
| Replace Euclidean distance with a learned Deep Bregman divergence under contrastive loss | Deep metric learning · Fashion MNIST | k-nn classification accuracy: 93.57 -> 94.00 | Helped |
| Replace Euclidean distance with a learned Deep Bregman divergence under contrastive loss | Deep metric learning · Cifar10 | k-nn classification accuracy: 79.40 -> 80.80 | Helped |
| Replace Euclidean distance with a learned Deep Bregman divergence under contrastive loss | Deep metric learning · SVHN | k-nn classification accuracy: 94.88 -> 94.12 | Hurt |
| Replace Euclidean distance with a learned Deep Bregman divergence under contrastive loss | Deep metric learning · STL10 | k-nn classification accuracy: 63.10 -> 60.91 | Hurt |
| Use distributional moment matching rather than treating generated points independently | Semi-supervised distributional clustering · Synthetic Gaussian clustering test set | RI mean: 0.639 -> 0.999 | Helped |
| Use distributional moment matching rather than treating generated points independently | Semi-supervised distributional clustering · Synthetic Gaussian clustering test set | ARI mean: 0.198 -> 0.997 | Helped |
| Use distributional moment matching rather than treating generated points independently | Semi-supervised distributional clustering · WHARF | ARI mean: 0.098 -> 0.364 | Helped |
Closest papers by meaning
- Learning to Approximate a Bregman Divergence (NeurIPS 2020 · similarity 0.93)
- Neural Bregman Divergences for Distance Learning (ICLR 2023 · similarity 0.93)
- $f$-Divergence Based Classification: Beyond the Use of Cross-Entropy (ICML 2024 · similarity 0.86)
- Learning Bregman Divergences with Application to Robustness (NeurIPS 2024 · similarity 0.86)
- Deep Disentangled Metric Learning (AAAI 2025 · similarity 0.85)
- Towards Improved Proxy-Based Deep Metric Learning via Data-Augmented Domain Adaptation (AAAI 2024 · 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.