Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Experiment Videos

Enhanced low-rank representation via sparse manifold adaption for semi-supervised learning.

Yong Peng1, Bao-Liang Lu2, Suhang Wang3

  • 1Center for Brain-like Computing and Machine Intelligence, Department of Computer Science and Engineering, Shanghai Jiao Tong University, Shanghai 200240, China.

Neural Networks : the Official Journal of the International Neural Network Society
|January 31, 2015
PubMed
Summary

Related Concept Videos

State Space Representation01:27

State Space Representation

744
The frequency-domain technique, commonly used in analyzing and designing feedback control systems, is effective for linear, time-invariant systems. However, it falls short when dealing with nonlinear, time-varying, and multiple-input multiple-output systems. The time-domain or state-space approach addresses these limitations by utilizing state variables to construct simultaneous, first-order differential equations, known as state equations, for an nth-order system.
Consider an RLC circuit, a...
744

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

Antagonistic effects during phenanthrene and pyrene co-degradation by Pseudoxanthomonas: Pathway competition and substrate preference.

Journal of hazardous materials·2026
Same author

A Roadmap to Navigate the Future of Neural Engineering.

Journal of neural engineering·2026
Same author

Enhance graph alignment for large language models.

Neural networks : the official journal of the International Neural Network Society·2026
Same author

ROS-responsive hydrogel loaded with capsaicin promotes tenogenic differentiation of tendon stem/progenitor cells and enhances tendon injury repair.

Materials today. Bio·2026
Same author

Cross-Stimulus Transfer Learning: Enhancing Emotion Recognition from Visual-Auditory to Olfactory Perception.

Annual International Conference of the IEEE Engineering in Medicine and Biology Society. IEEE Engineering in Medicine and Biology Society. Annual International Conference·2025
Same author

Enhancing Depression Detection from Emotion EEG with Temporal-Spatial-Spectral Representation Learning.

Annual International Conference of the IEEE Engineering in Medicine and Biology Society. IEEE Engineering in Medicine and Biology Society. Annual International Conference·2025

This study introduces Manifold Low-Rank Representation (MLRR) for improved graph construction in pattern recognition. MLRR enhances low-rank representation by incorporating local data manifold structures for better classification and clustering.

Area of Science:

  • Pattern Recognition
  • Machine Learning
  • Data Mining

Background:

  • Graph construction is crucial for pattern recognition tasks like clustering and classification.
  • Low-Rank Representation (LRR) is a prominent graph-based learning model used in spectral clustering and semi-supervised learning (SSL).
  • Existing LRR methods often overlook the local geometrical structure of data, limiting their discriminative power.

Purpose of the Study:

  • To propose an enhanced Low-Rank Representation (LRR) method that integrates local data manifold structures.
  • To develop Manifold Low-Rank Representation (MLRR) for learning robust low-rank data representations.
  • To improve graph construction for semi-supervised learning (SSL) by considering both global and local data properties.

Main Methods:

Keywords:
Face recognitionGraph constructionLow-rank representationSemi-supervised learningSparse manifold adaption

Related Experiment Videos

  • Proposed Manifold Low-Rank Representation (MLRR) by adapting sparse manifold learning.
  • Identified local data manifold structure using geometric sparsity and sparse representation.
  • Incorporated a regularizer into LRR to preserve geometric constraints, combining global LRR with local manifold information.

    • Main Results:

    • MLRR effectively learns low-rank data representations by considering local manifold structures.
    • The proposed method integrates global information (low-rank property) with local information (manifold structure).
    • Experimental results on semi-supervised classification tasks show MLRR outperforms existing state-of-the-art graph construction approaches.

    Conclusions:

    • MLRR offers a significant advancement in graph-based learning by incorporating manifold structures.
    • The method enhances discriminative capabilities in pattern recognition tasks, particularly SSL.
    • MLRR provides a superior approach to graph construction compared to traditional LRR methods.