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

Information discriminant analysis: feature extraction with an information-theoretic objective.

Zoran Nenadic1

  • 1Department of Biomedical Engineering, University of California, Irvine, 3120 Natural Sciences II, Irvine, CA 92697-2715, USA. znenadic@uci.edu

IEEE Transactions on Pattern Analysis and Machine Intelligence
|June 15, 2007
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

Perception of brain-computer interface implantation surgery for motor, sensory, and autonomic restoration in spinal cord injury and stroke.

Frontiers in neuroscience·2026
Same author

Real-time brain-computer interface control of walking exoskeleton with bilateral sensory feedback.

Brain stimulation·2026
Same author

Robotic Ankle Assessment Post-Stroke: Reliability, Comparison to Therapists, and Benchmark Dataset Development.

Sensors (Basel, Switzerland)·2025
Same author

The Weak Relationship Between Ankle Proprioception and Gait Speed After Stroke: A Robotic Assessment Study.

Neurorehabilitation and neural repair·2025
Same author

The Relationship Between Stroke-Related Cognitive Impairment and Robotic Proprioceptive Assessment.

IEEE ... International Conference on Rehabilitation Robotics : [proceedings]·2025
Same author

Using a Robotic Device to Quantify the Impact of Conventional Gait Training on Ankle Proprioception After Stroke.

IEEE ... International Conference on Rehabilitation Robotics : [proceedings]·2025
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

A new information-theoretic method transforms observation data into a low-dimensional feature subspace for classification. This technique optimizes an objective function, offering advantages over existing methods and demonstrating Bayes-optimality for improved classification performance.

Area of Science:

  • Machine Learning
  • Information Theory
  • Pattern Recognition

Background:

  • Feature extraction is crucial for effective classification.
  • Existing linear transformation methods have limitations.
  • Information-theoretic approaches offer potential for optimal feature representation.

Purpose of the Study:

  • To develop a novel linear transformation technique for classification.
  • To optimize an information-theoretic objective function for feature subspace creation.
  • To compare the proposed method with existing linear discriminant analysis techniques.

Main Methods:

  • Utilized elementary information-theoretic tools.
  • Developed a novel objective function for numerical optimization.
  • Analyzed the properties of the objective function concerning mutual information and Bayes error.

Related Experiment Videos

  • Investigated conditions for Bayes-optimality and computational acceleration.

    • Main Results:

    • The proposed method achieves linear transformation into a low-dimensional feature subspace.
    • The novel objective function exhibits desirable properties similar to mutual information and Bayes error.
    • The method demonstrates favorable performance compared to linear discriminant analysis on various datasets.
    • Computational acceleration techniques were developed for feasible solutions.

    Conclusions:

    • The novel information-theoretic technique provides an effective approach for feature extraction in classification.
    • The method shows promise for achieving Bayes-optimal classification.
    • The technique offers advantages over existing linear discriminant-based feature extraction methods.