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

Subclass discriminant analysis.

Manli Zhu1, Aleix M Martinez

  • 1Department of Electrical and Computer Engineering, The Ohio State University, 205 Dreese Lab, 2015 Neil Ave., Columbus, OH 43210, USA. zhum@ece.osu.edu

IEEE Transactions on Pattern Analysis and Machine Intelligence
|August 5, 2006
PubMed
Summary
This summary is machine-generated.

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

Current Status and Future Prospects of Research on Sepsis-Related Acute Kidney Injury.

International journal of molecular sciences·2026
Same author

Statistical modeling of immunoprecipitation efficiency of MeRIP-seq data enabled accurate detection and quantification of epitranscriptome.

Computational and structural biotechnology journal·2025
Same author

Evaluation of the nutritional, <i>in vitro</i> protein digestive and bioactive characteristics of a quinoa-based protein beverage.

Food chemistry: X·2025
Same author

Ultrahigh Exchange Bias Field/Coercive Field Ratio in In Situ Formed Two-Dimensional Magnetic Te-Cr<sub>2</sub>O<sub>3</sub>/Cr<sub>5</sub>Te<sub>6</sub> Heterostructures.

Advanced materials (Deerfield Beach, Fla.)·2025
Same author

Association of Circulating Homocysteine Level with the Risk of Nephropathy in Type 2 Diabetes Mellitus: A Meta-Analysis.

Hormone and metabolic research = Hormon- und Stoffwechselforschung = Hormones et metabolisme·2024
Same author

Predictive models of sepsis-associated acute kidney injury based on machine learning: a scoping review.

Renal failure·2024
Same journal

Relation DETR+: Exploring Explicit Position Relation Prior for Dense Prediction.

IEEE transactions on pattern analysis and machine intelligence·2026
Same journal

RBF++: Quantifying and Optimizing Reasoning Boundaries across Measurable and Unmeasurable Capabilities for Chain-of-Thought Reasoning.

IEEE transactions on pattern analysis and machine intelligence·2026
Same journal

CAFE: Cross-View Adaptive Fusion and Cluster Center Enhancement for Robust Multi-View Clustering.

IEEE transactions on pattern analysis and machine intelligence·2026
Same journal

DIVER: Reinforced Diffusion Breaks Imitation Bottlenecks in End-to-End Autonomous Driving.

IEEE transactions on pattern analysis and machine intelligence·2026
Same journal

Ethics-Aware Safe Reinforcement Learning for Rare-Event Risk Control in Interactive Urban Driving.

IEEE transactions on pattern analysis and machine intelligence·2026
Same journal

Learning Shape Anchors for Holistic Indoor Scene Understanding.

IEEE transactions on pattern analysis and machine intelligence·2026
See all related articles

This study introduces a new method for Discriminant Analysis (DA) using Gaussian mixture models to handle unknown data distributions. The approach effectively determines the optimal number of subclasses, outperforming existing DA algorithms in experiments.

Area of Science:

  • Machine Learning
  • Statistical Pattern Recognition
  • Data Mining

Background:

  • Discriminant Analysis (DA) algorithms are widely used for high-dimensional data but are often tuned to specific, unknown distributions.
  • Current methods require trial-and-error selection of the best-fitting DA algorithm for a given problem.
  • A universal approach is needed to accommodate various data distributions without prior knowledge.

Purpose of the Study:

  • To develop a single Discriminant Analysis formulation applicable to diverse data distributions.
  • To address the challenge of determining the optimal number of Gaussian components (subclasses) per class in mixture models.
  • To introduce novel criteria for optimally dividing each class into subclasses.

Main Methods:

  • Approximating unknown class probability density functions (pdfs) using mixtures of Gaussians.

Related Experiment Videos

  • Deriving two criteria to determine the optimal number of Gaussians per class.
  • Conducting extensive experiments on five databases to validate the proposed method.

    • Main Results:

    • The proposed method, utilizing Gaussian mixture models, demonstrated superior or comparable performance against established algorithms.
    • Experimental results showed the effectiveness of the derived criteria in optimizing subclass determination.
    • The method proved robust across various datasets, handling unknown distributions effectively.

    Conclusions:

    • The developed criteria for Gaussian mixture modeling offer an effective solution for Discriminant Analysis with unknown distributions.
    • This approach provides a more universal and efficient alternative to traditional, distribution-specific DA methods.
    • The method consistently achieves top-tier performance, making it a valuable tool for high-dimensional data analysis.