论文阅读 [TPAMI-2022] Truncated Robust Principle Component Analysis With A General Optimization Framewor

论文阅读 [TPAMI-2022] Truncated Robust Principle Component Analysis With A General Optimization Framework

论文搜索(studyai.com)

搜索论文: Truncated Robust Principle Component Analysis With A General Optimization Framework

搜索论文: http://www.studyai.com/search/whole-site/?q=Truncated+Robust+Principle+Component+Analysis+With+A+General+Optimization+Framework

关键字(Keywords)

Robustness; Principal component analysis; Analytical models; Optimization; Image reconstruction; Data models; Adaptive optics; Robust principle component analysis (RPCA); unsupervised dimensionality reduction; truncated loss; non-convex optimization

机器学习; 机器视觉; 运筹与优化

无监督学习; 凸优化; 非凸问题; 维数约简; 主成分分析; 图像重建

摘要(Abstract)

Recently, several robust principle component analysis (RPCA) models have been proposed to improve the robustness of principle component analysis (PCA).

最近，为了提高主成分分析（PCA）的鲁棒性，人们提出了几种鲁棒主成分分析（RPCA）模型。.

But an important problem that the robustness to outliers affects the discrimination of correct samples has not been solved yet.

但是，对异常值的鲁棒性影响正确样本的识别这一重要问题尚未得到解决。.

To solve this problem, we propose a truncated robust principle component analysis (T-RPCA) model which treats correct samples and outliers separately.

为了解决这个问题，我们提出了一个截断稳健主成分分析（T-RPCA）模型，该模型分别处理正确样本和异常值。.

In fact, the proposed model performs an implicitly truncated weighted learning scheme which is more reasonable for robustness learning respective to previous works.

事实上，所提出的模型执行了一种隐式截断加权学习方案，与以前的工作相比，该方案对于鲁棒性学习更为合理。.

Moreover, we propose a re-weighted (RW) optimization framework to solve a general problem and generalize two sub-frameworks upon it.

此外，我们还提出了一个重新加权（RW）优化框架来解决一般问题，并在此基础上推广了两个子框架。.

To be specific, the first sub-framework orients a general truncated loss optimization problem which contains the objective problem of T-RPCA, and the second one focuses on a general singular-value based optimization problem.

具体来说，第一个子框架面向一个包含T-RPCA目标问题的一般截断损耗优化问题，第二个子框架关注一个基于奇异值的一般优化问题。.

Besides, we provide rigorously theoretical guarantees for the proposed model, RW framework and sub-frameworks.

此外，我们还为所提出的模型、RW框架和子框架提供了严格的理论保证。.

Empirical studies demonstrate that the proposed T-RPCA model outperforms previous RPCA models on reconstruction and classification tasks…

实证研究表明，提出的T-RPCA模型在重建和分类任务上优于以前的RPCA模型。。.

作者(Authors)

[‘Feiping Nie’, ‘Danyang Wu’, ‘Rong Wang’, ‘Xuelong Li’]