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

Feature extraction using information-theoretic learning.

Kenneth E Hild1, Deniz Erdogmus, Kari Torkkola

  • 1Biomagnetic Imaging Laboratory, University of California at San Francisco, Room C-324B, San Francisco, CA 94122, USA. k.hild@ieee.org

IEEE Transactions on Pattern Analysis and Machine Intelligence
|August 26, 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

Human-AI Cooperation in Healthcare and Rehabilitation.

Delaware journal of public health·2026
Same author

Temporal point process modeling of aggressive behavior onset in psychiatric inpatient youths with autism.

Scientific reports·2026
Same author

Training sparse convolutional deep predictive coding networks with attention.

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

Corticomorphic Hybrid CNN-SNN Architecture for EEG-Based Low-Footprint Low-Latency Auditory Attention Detection.

Annals of biomedical engineering·2026
Same author

Deep Learning-Based Prediction of Cardiopulmonary Disease in Retinal Images of Premature Infants.

JAMA ophthalmology·2026
Same author

A generative spike prediction model using behavioral reinforcement for re-establishing neural functional connectivity.

Nature computational science·2026
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 training feature extractors independently, achieving performance comparable to simultaneous training. This approach offers implementation advantages while maintaining classification accuracy.

Area of Science:

  • Machine Learning
  • Information Theory
  • Pattern Recognition

Background:

  • Classification systems comprise feature extractors and classifiers, trainable independently or simultaneously.
  • Independent training offers implementation flexibility, while simultaneous training can directly minimize classification error.
  • Existing criteria like Minimum Classification Error favor simultaneous training, while Mutual Information allows for both approaches.

Purpose of the Study:

  • To introduce and evaluate an information-theoretic criterion for independent feature extractor training.
  • To maximize the mutual information between class labels and feature extractor output using Renyi's entropy.
  • To demonstrate that independent training can achieve performance on par with simultaneous training methods.

Main Methods:

Related Experiment Videos

  • Developed a novel information-theoretic criterion for feature extractor training.
  • Employed nonparametric estimation of Renyi's entropy.
  • Maximized an approximation of mutual information between class labels and extractor outputs.

Main Results:

  • The proposed independent training method achieved performance comparable to simultaneous training.
  • Evaluated against three simultaneous feature extraction methods, the proposed approach showed equivalent or superior results.
  • Demonstrated the efficacy of using mutual information for independent feature extractor optimization.

Conclusions:

  • Independent training of feature extractors using the proposed information-theoretic criterion is a viable and effective approach.
  • This method offers practical implementation benefits without compromising classification performance.
  • The study highlights the potential of information-theoretic criteria for optimizing feature extraction in classification systems.