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

Combinative neural network and its applications.

Yaqiu Chen1, Shangxu Hu, Dezhao Chen

  • 1Department of Chemical Engineering, National Key Laboratory of Industrial Control Technology, Zhejiang University, 310027, Hangzhou, People's Republic of China. sxhu@mail.hz.zj.cn

Computational Biology and Chemistry
|August 21, 2003
PubMed
Summary

A novel combinative neural network (CN) integrates partial least squares (PLS) to enhance multi-layered feed forward neural networks (MLFF NN). This approach effectively handles non-linear data and prevents overfitting, offering a time-saving solution for complex mapping and classification tasks.

Related Concept Videos

You might also read

Related Articles

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

Sort by
Same author

RHEBL1 expression is activated by the Oct4-Sox2 complex in oral squamous cell carcinoma.

Functional & integrative genomics·2026
Same author

Exploring the celecoxib-cervical cancer relationship using <i>in vitro</i>, network pharmacology, and Mendelian randomization approaches.

Frontiers in medicine·2026
Same author

PTN/IGF-2 signaling modulates endometrial decidualization and immune cell trafficking to facilitate pregnancy maintenance.

Frontiers in immunology·2026
Same author

Manganese Biomineralized Ferritin Nanoplatforms with Shielding and Stimuli-Responsive Release for Potentiated Ferroptosis and Multimodal Ovarian Cancer Therapy.

International journal of nanomedicine·2026
Same author

Polyvinyl chloride microplastics exposure accelerates endometrial cancer progression via regulating AHR/CYP1A1 signaling pathway.

Toxicology·2026
Same author

DNA nanotherapeutics facilitate efficient gynecological cancers treatment.

Journal of controlled release : official journal of the Controlled Release Society·2025

Area of Science:

  • Artificial Intelligence
  • Machine Learning
  • Computational Science

Background:

  • Multi-layered feed forward neural networks (MLFF NN) are widely used for complex data analysis.
  • Traditional MLFF NNs can struggle with non-linear relationships and are prone to overfitting.
  • Partial least squares (PLS) analysis offers methods for variable correlation and dimensionality reduction.

Purpose of the Study:

  • To introduce a novel combinative neural network (CN) architecture.
  • To leverage PLS analysis to improve the hidden layer of MLFF NNs.
  • To enhance the performance of neural networks in non-linear mapping and classification tasks.

Main Methods:

  • A combinative neural network (CN) was developed by incorporating PLS analysis into the hidden layer of MLFF NNs.

Related Experiment Videos

  • PLS was utilized to reorganize hidden node outputs, circumventing variable correlation.
  • The method was validated on both a non-linear approximation problem and a non-linear pattern classification problem.

    • Main Results:

    • The proposed CN effectively addressed non-linear relationships between input and output data.
    • PLS integration in the CN successfully circumvented variable correlation and mitigated overfitting.
    • Performance comparisons with conventional MLFF NNs demonstrated superior results for the CN approach.

    Conclusions:

    • The combinative neural network (CN) provides a robust method for non-linear mapping and classification.
    • The integration of PLS analysis offers significant advantages in handling complex datasets.
    • The CN approach presents a time-saving and effective alternative to traditional neural network models.