Meta-learning在低熵环境下更优，提出MINO框架提升无监督少样本性能。
Is Meta-Learning Out? Rethinking Unsupervised Few-Shot Classification with Limited Entropy
- 用受限熵设置公平对比，发现元学习泛化能力更强
- 在标签噪声和任务异构下表现更鲁棒，适合无监督场景
- 提出MINO框架，结合DBSCAN与动态头，提升无监督少样本效果
元学习是解决少样本任务的强大范式。然而近期研究表明，采用全类训练策略的模型在少样本分类任务中可达到与元学习相当的性能。为验证元学习的价值，我们建立了一个熵受限的监督设置以实现公平比较。通过理论分析与实验验证，我们发现元学习具有更紧的泛化界。研究揭示：在有限熵条件下，元学习更高效，对标签噪声和异构任务更具鲁棒性，因此特别适合无监督任务。基于此，我们提出MINO——一种专为提升无监督性能设计的元学习框架。MINO利用自适应聚类算法DBSCAN构建无监督任务，结合动态头与基于稳定性的元缩放器，增强对标签噪声的鲁棒性。大量实验验证了其在多个无监督少样本及零样本任务中的有效性。
原文摘要 · Abstract (English)
Meta-learning is a powerful paradigm for tackling few-shot tasks. However, recent studies indicate that models trained with the whole-class training strategy can achieve comparable performance to those trained with meta-learning in few-shot classification tasks. To demonstrate the value of meta-learning, we establish an entropy-limited supervised setting for fair comparisons. Through both theoretical analysis and experimental validation, we establish that meta-learning has a tighter generalization bound compared to whole-class training. We unravel that meta-learning is more efficient with limited entropy and is more robust to label noise and heterogeneous tasks, making it well-suited for unsupervised tasks. Based on these insights, We propose MINO, a meta-learning framework designed to enhance unsupervised performance. MINO utilizes the adaptive clustering algorithm DBSCAN with a dynamic head for unsupervised task construction and a stability-based meta-scaler for robustness against label noise. Extensive experiments confirm its effectiveness in multiple unsupervised few-shot and zero-shot tasks.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。