arXiv：cs.LG（机器学习，全量分类）· Rim Hajal, Mathieu Besan\c{c}on, J\'er\^ome Malick·· 1 天前AI 评分31
基于熵最优传输的低预算主动学习
Low-Budget Active Learning through Entropic Optimal Transport
AI 导读
研究者提出用熵最优传输中的 Sinkhorn 散度作为 coreset 选择准则，在预训练自监督模型提取的特征空间中直接进行低预算主动学习。该方法可获得与维度无关的样本复杂度结果，并支持高效梯度计算，配合基于交换的局部搜索快速求解。在图像基准和医学数据集上，其低预算表现优于现有启发式方法。
正文
Abstract:We consider low-budget active learning, which consists of selecting a limited number of points, the coreset, such that a model can be trained to high accuracy on the selection only. This problem is particularly relevant in contexts where labeling requires costly expert intervention, as in medical applications. We leverage features extracted from a pretrained self-supervised model to represent the data, and perform coreset selection directly in this feature space. In this paper, we use entropic optimal transport, specifically the Sinkhorn divergence, as the coreset selection criterion, which first allows us to get dimension-free sample complexity results, and second admits computationally efficient gradient evaluations. This opens the way to using gradient-based algorithms to rapidly compute solution candidates, further improved by a swap-based local search, with guarantees on the solution quality. Experiments on image benchmarks and medical datasets show that our method outperforms state-of-the-art heuristics in low-budget settings.
| Subjects: | Machine Learning (cs.LG); Optimization and Control (math.OC) |
| Cite as: | arXiv:2610.01199 [cs.LG] |
| (or arXiv:2610.01199v1 [cs.LG] for this version) | |
| https://doi.org/10.48550/arXiv.2610.01199 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Rim Hajal [view email]
[v1]
Thu, 1 Oct 2026 07:11:24 UTC (55 KB)
来源：arXiv：cs.LG（机器学习，全量分类） · arxiv.org