arXiv 2603.157082026-03-16cs.LGcs.AI

掌控少数类：一种基于不确定性的多专家框架用于挑战性尾部序列学习

Mastering the Minority: An Uncertainty-guided Multi-Expert Framework for Challenging-tailed Sequence Learning

Key Laboratory of Cyberspace Big Data Intelligent Security, Ministry of Education（教育部空天地大数据智能安全重点实验室）；School of Artificial Intelligence, Chongqing University of Post and Telecommunications（重庆邮电大学人工智能学院）；School of Computer Science and Technology, East China Normal University（华东师范大学计算机科学与技术学院）；Chongqing Key Laboratory of Brain-Inspired Cognitive Computing and Educational Rehabilitation for Children with Special Needs, Chongqing Normal University（重庆市脑科学与认知计算及特殊儿童教育康复重点实验室）

Ye Wang, Zixuan Wu, Lifeng Shen, Jiang Xie, Xiaoling Wang, Hong Yu, Guoyin Wang

阅读 PDF

arXiv 页面 ↗HTML 原文 ↗

AI总结：

本文提出基于不确定性的多专家融合网络（UME）框架，通过参数高效建模、序列专业化和不确定性引导融合机制，解决序列学习中少数类识别难题，实验表明在四个公开数据集上取得最佳性能。

摘要：

不平衡数据分布仍然是序列学习中的关键挑战，导致模型容易识别频繁类别而无法有效检测少数类。混合专家模型提供了一种可扩展的解决方案，但其应用常受参数效率低、专家专业化差和预测冲突解决困难的限制。为有效掌控少数类，我们提出不确定性多专家融合网络（UME）框架。UME具有三个核心创新：首先，我们采用Ensemble LoRA进行参数高效建模，显著减少可训练参数数量。其次，我们引入由Dempster-Shafer理论（DST）引导的序列专业化，确保在挑战性尾部类别上的有效专业化。最后，不确定性引导融合机制利用DST的置信度度量动态加权专家意见，通过优先考虑最自信的专家解决冲突，实现可靠最终预测。在四个公开的层次化文本分类数据集上进行的广泛实验表明，UME在性能上达到最佳。我们在单个类别上比最佳基线提高了高达17.97%，同时将可训练参数减少了高达10.32%。研究结果表明，基于不确定性的专家协调是一种解决挑战性尾部序列学习的系统性策略。我们的代码可在https://github.com/CQUPTWZX/Multi-experts获取。

英文摘要：

Imbalanced data distribution remains a critical challenge in sequential learning, leading models to easily recognize frequent categories while failing to detect minority classes adequately. The Mixture-of-Experts model offers a scalable solution, yet its application is often hindered by parameter inefficiency, poor expert specialization, and difficulty in resolving prediction conflicts. To Master the Minority classes effectively, we propose the Uncertainty-based Multi-Expert fusion network (UME) framework. UME is designed with three core innovations: First, we employ Ensemble LoRA for parameter-efficient modeling, significantly reducing the trainable parameter count. Second, we introduce Sequential Specialization guided by Dempster-Shafer Theory (DST), which ensures effective specialization on the challenging-tailed classes. Finally, an Uncertainty-Guided Fusion mechanism uses DST's certainty measures to dynamically weigh expert opinions, resolving conflicts by prioritizing the most confident expert for reliable final predictions. Extensive experiments across four public hierarchical text classification datasets demonstrate that UME achieves state-of-the-art performance. We achieve a performance gain of up to 17.97\% over the best baseline on individual categories, while reducing trainable parameters by up to 10.32\%. The findings highlight that uncertainty-guided expert coordination is a principled strategy for addressing challenging-tailed sequence learning. Our code is available at https://github.com/CQUPTWZX/Multi-experts.

机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。