English(EN) Constructing Structured Decision Sources for Consensus-Based Pseudo-Label Learning

新方法构建多样化决策源以改进伪标签学习

作者 PulseAugur 编辑部 · [1 个来源] · 2026-10-09 04:00

研究人员开发了一种通过构建更具信息量的决策源来改进伪标签学习的方法。该方法不依赖多个相同的模型，而是修改共享图表示的内部结构以创建多样化的证据。通过选择这些构建的源的互补子集，该方法提高了伪标签的精度，从而在 Cora、CiteSeer 和 PubMed 等数据集上实现了具有竞争力的下游准确性。这项工作强调了在基于共识的伪标签学习中，构建源比简单地增加模型数量更重要。 AI

影响这项研究通过提高伪标签的质量，有望实现更高效、更准确的机器学习模型训练。

排序理由详细介绍一种改进伪标签学习新方法的学术论文。[lever_c_demoted from research: ic=1 ai=1.0]

在 arXiv cs.LG 阅读 →

AI 生成摘要 · Google Gemini · 来自 1 个来源。我们如何撰写摘要 →

本文如何被排名

Signal score: 13 / 100; Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket: Tool; 详细介绍一种改进伪标签学习新方法的学术论文。[lever_c_demoted from research: ic=1 ai=1.0]
Source corroboration: Single-source cluster; Only one publisher covered this so far. Single-source stories can still rank when the publisher is high-authority, but they lack cross-source corroboration.
Topics: paper, other; Editorial topic classification. Feeds into how the story surfaces on /topic/<slug> hub pages and into the per-entity coverage mix.
AI-industry relevance: High; Clearly on-topic for AI-industry coverage.
Story freshness: Breaking (< 6h); Fresh story with cross-source coverage still developing. Ranking may shift as more sources report.

完整方法见我们的编辑标准。

报道来源 [1]

arXiv cs.LG TIER_1 English(EN) · Long Wang · 2026-10-09 04:00

构建用于共识伪标签学习的结构化决策源

arXiv:2610.11621v1 Announce Type: new Abstract: Consensus can make pseudo-label learning more reliable, but only when its predictors contribute genuinely different evidence. Multiple models that repeat the same boundary provide additional votes without additional information. We …

本文如何被排名

报道来源 [1]

构建用于共识伪标签学习的结构化决策源

相关实体

相关话题