Research map / Learning theory and algorithms / Prediction and decision losses
Simplex Clustering via sBeta With Applications to Online Adjustment of Black-Box Predictions
Florent Chiaroni, Malik Boudiaf, Amar Mitiche, Ismail Ben Ayed
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 |
|---|---|---|---|
| add sBeta scaling (delta=0.15) to Beta-based clustering (ablation) | k-Betas · VISDA-C | accuracy: 19.8 -> 56.0 | Improved |
| add sBeta scaling (delta=0.15) to Beta-based clustering (ablation) | k-Betas · SVHN->MNIST | accuracy: 51.5 -> 76.6 | Improved |
| add sBeta scaling on synthetic Dirichlet data (ablation) | k-Betas · Simu | NMI: 81.1 -> 79.2 | Hurt |
| use vertex init instead of k-means++ (ablation) | k-sBetas · SVHN->MNIST | accuracy: 69.8 -> 76.6 | Improved |
| use vertex init instead of k-means++ (ablation) | k-sBetas · VISDA-C | accuracy: 47.2 -> 56.0 | Improved |
| use vertex init instead of k-means++ (ablation) | k-means · SVHN->MNIST | accuracy: 66.6 -> 68.9 | Improved |
| use vertex init instead of k-means++ (ablation) | k-medoids · SVHN->MNIST | accuracy: 51.9 -> 71.3 | Improved |
| cluster simplex softmax instead of argmax | k-sBetas · VISDA-C | accuracy: 53.1 -> 56.0 | Improved |
| replace Euclidean k-means with k-sBetas on simplex | k-means · VISDA-C | accuracy: 47.9 -> 56.0 | Improved |
| cluster simplex predictions instead of feature maps with k-means | k-means · VISDA-C | accuracy: 22.7 -> 47.9 | Improved |
| downsample prediction map by factor 8 before clustering (ablation) | k-sBetas · GTA5 road->Cityscapes road | mIoU: 65.7 -> 65.7 | No clear effect |
| use method of moments instead of MLE-1000 for parameter estimation (ablation) | k-sBetas · SVHN->MNIST | running time (s): 64.32 -> 0.48 | Improved |
Closest papers by meaning
- Deep Clustering: On the Link Between Discriminative Models and K-Means (TPAMI 2019 · similarity 0.86)
- Towards K-means-friendly Spaces: Simultaneous Deep Learning and Clustering (ICML 2017 · similarity 0.86)
- Generalised Mutual Information for Discriminative Clustering (NeurIPS 2022 · similarity 0.84)
- Samples Are Not Equal: A Sample Selection Approach for Deep Clustering (ICLR 2026 · similarity 0.84)
- Deep Semantic Clustering by Partition Confidence Maximisation (CVPR 2020 · similarity 0.84)
- Improved Learning-augmented Algorithms for k-means and k-medians Clustering (ICLR 2023 · 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.
