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

Optimized multilayer perceptrons for molecular classification and diagnosis using genomic data.

Zuyi Wang1, Yue Wang, Jianhua Xuan

  • 1Center for Genetic Medicine, Children's National Medical Center Washington, DC 20010, USA.

Bioinformatics (Oxford, England)
|January 13, 2006
PubMed
Summary

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

MAE-UNETR++: Masked Autoencoder Pretraining for 3-D Lung Nodule Segmentation.

bioRxiv : the preprint server for biology·2026
Same author

Preclinical efficacy of a gene therapy for <i>CHKB</i>-mediated muscular dystrophy.

Molecular therapy. Advances·2026
Same author

Vamorolone for Duchenne Muscular Dystrophy: A Cross-Trial Efficacy Comparison With Classic Corticosteroids From the FOR-DMD Trial.

Neurology·2026
Same author

Vamorolone Safety, Pharmacokinetics, and Exploratory Efficacy in Duchenne Muscular Dystrophy: A Phase II, Nonrandomized, Multiple-Dose Study in 2-<4-Year-Old Boys.

Neurology·2026
Same author

Results of a phase II open-label, multiple-dose study of vamorolone (VBP15-006) in 7- to < 18-year-old boys with duchenne muscular dystrophy.

Journal of neurology·2026
Same author

Exposure-response of serum biomarkers to vamorolone, a dissociative corticosteroidal anti-inflammatory drug, in 4- to <7-year children.

Steroids·2025

This study introduces an optimized Multilayer Perceptron (MLP) for genomic data analysis, improving diagnostic classification accuracy by addressing the curse of dimensionality and overfitting. The new method enhances prediction performance compared to standard MLPs.

Area of Science:

  • Genomics
  • Machine Learning
  • Bioinformatics

Background:

  • Multilayer perceptrons (MLPs) are effective for diagnostic classification using high-dimensional genomic data.
  • High dimensionality in genomic datasets often leads to performance degradation and overfitting in MLPs.
  • Existing MLP methods may struggle to achieve acceptable prediction accuracy with large genomic datasets.

Purpose of the Study:

  • To design and implement an MLP optimization scheme for improved diagnostic classification of genomic data.
  • To address the challenges of high dimensionality and overfitting in MLPs applied to genomic data.
  • To enhance the prediction accuracy and performance of MLPs in large microarray datasets.

Main Methods:

  • Developed an MLP optimization scheme based on Fisher linear discriminant analysis.

Related Experiment Videos

  • Optimized MLP parameters and architecture for a two-layer MLP.
  • Evaluated the optimized MLP on large microarray datasets.

    • Main Results:

    • The optimized MLP effectively mitigated the curse of dimensionality in large microarray datasets.
    • Significant improvements were observed in Bayes classification accuracy compared to conventional MLPs.
    • Enhanced convergence properties and area under the receiver operating characteristic curve (A(z)) were achieved.

    Conclusions:

    • The proposed MLP optimization scheme offers superior performance for genomic data classification.
    • This method effectively overcomes limitations of standard MLPs in high-dimensional genomic analyses.
    • The optimized MLP provides a more accurate and reliable tool for genomic diagnostic classification.