arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2201.11231cs.LGcs.AI
收藏

知识共享和转移的差距最小化

Gap Minimization for Knowledge Sharing and Transfer

University of Western Ontario(西安大略大学);Massachusetts Institute of Technology(麻省理工学院);Université Laval(拉瓦尔大学);Beihang University(北京航空航天大学);McGill University(麦吉尔大学);University of Pennsylvania(宾夕法尼亚大学)

Boyu Wang, Jorge Mendez, Changjian Shui, Fan Zhou, Di Wu, Gezheng Xu, Christian Gagné, Eric Eaton

摘要:

在过去二十年中,通过知识共享和转让从多个相关任务中学习变得越来越重要。为了成功地将信息从一个任务转移到另一个任务,理解域之间的相似性和差异性至关重要。在本文中,我们介绍了\emph{performance gap}的概念,一个直观的和新颖的测量学习任务之间的距离。与用作约束任务之间预期风险差异的工具的现有措施不同(例如,$\mathcal{H}$-发散或差异距离),我们从理论上表明,性能差距可以被视为一个数据和算法相关的正则化,控制模型的复杂性,并导致更好的保证。更重要的是,它还提供了新的见解,并激发了一个新的原则,设计战略的知识共享和转让:差距最小化。我们实例化这一原则与两个算法:1。gapBoost,一种新颖且有原则的提升算法,其显式地最小化用于迁移学习的源域与目标域之间的性能差距;以及2. gapMTNN是一种表示学习算法,它将间隙最小化重新表述为多任务学习的语义条件匹配。我们对迁移学习和多任务学习基准数据集的广泛评估表明,我们的方法优于现有的基线。

英文摘要:

Learning from multiple related tasks by knowledge sharing and transfer has become increasingly relevant over the last two decades. In order to successfully transfer information from one task to another, it is critical to understand the similarities and differences between the domains. In this paper, we introduce the notion of \emph{performance gap}, an intuitive and novel measure of the distance between learning tasks. Unlike existing measures which are used as tools to bound the difference of expected risks between tasks (e.g., $\mathcal{H}$-divergence or discrepancy distance), we theoretically show that the performance gap can be viewed as a data- and algorithm-dependent regularizer, which controls the model complexity and leads to finer guarantees. More importantly, it also provides new insights and motivates a novel principle for designing strategies for knowledge sharing and transfer: gap minimization. We instantiate this principle with two algorithms: 1. gapBoost, a novel and principled boosting algorithm that explicitly minimizes the performance gap between source and target domains for transfer learning; and 2. gapMTNN, a representation learning algorithm that reformulates gap minimization as semantic conditional matching for multitask learning. Our extensive evaluation on both transfer learning and multitask learning benchmark data sets shows that our methods outperform existing baselines.

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。

↑