arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2308.00142cs.LG
收藏

斯蒂费尔流形上的半监督拉普拉斯学习

Semi-Supervised Laplace Learning on Stiefel Manifolds

University of California San Diego(加利福尼亚大学圣迭戈分校);National Chung Hsing University(国立中兴大学)

Chester Holtz, Pengwen Chen, Alexander Cloninger, Chung-Kuan Cheng, Gal Mishne

AI总结:

针对低标签率下经典拉普拉斯学习算法的退化问题,提出基于信任域子问题非凸推广的半监督学习框架,结合正交Procrustes解与序列子空间优化,并设计新型中心性度量选择样本,在各标签率下分类误差均优于现有方法。

摘要:

受解决低标签率下经典拉普拉斯学习算法退化问题的需求驱动,我们提出将基于图的半监督学习重新表述为信任域子问题(Trust-Region Subproblem, TRS)的非凸推广。这一重构的动机是拉普拉斯特征向量在无限未标记数据极限下的适定性。为求解该问题,我们首先证明一阶条件意味着流形对齐问题的解,且经典正交 Procrustes 问题的解可用于高效找到便于进一步优化的优质分类器。为处理优化问题,我们为基于图的半监督学习(SSL)开发了序列子空间优化框架。接着,我们探讨了低标签率下有监督样本选择的关键性。我们利用从图拉普拉斯矩阵某一子矩阵的主特征向量推导得到的新型中心性度量,对信息性样本进行了表征。实验证明,在极低、中等和高标签率下,我们的框架相比近期最先进的及经典的半监督学习方法,均实现了更低的分类误差。

英文摘要:

Motivated by the need to address the degeneracy of canonical Laplace learning algorithms in low label rates, we propose to reformulate graph-based semi-supervised learning as a nonconvex generalization of a \emph{Trust-Region Subproblem} (TRS). This reformulation is motivated by the well-posedness of Laplacian eigenvectors in the limit of infinite unlabeled data. To solve this problem, we first show that a first-order condition implies the solution of a manifold alignment problem and that solutions to the classical \emph{Orthogonal Procrustes} problem can be used to efficiently find good classifiers that are amenable to further refinement. To tackle refinement, we develop the framework of Sequential Subspace Optimization for graph-based SSL. Next, we address the criticality of selecting supervised samples at low-label rates. We characterize informative samples with a novel measure of centrality derived from the principal eigenvectors of a certain submatrix of the graph Laplacian. We demonstrate that our framework achieves lower classification error compared to recent state-of-the-art and classical semi-supervised learning methods at extremely low, medium, and high label rates.

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。

↑