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 problem-dependent design for error-correcting output codes.

Sergio Escalera1, David M J Tax, Oriol Pujol

  • 1Computer Vision Center, Campus UAB, Bellaterra, Barcelona, Spain. sergio@maia.ub.es

IEEE Transactions on Pattern Analysis and Machine Intelligence
|April 19, 2008
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

From code to care in hours: Why "go fast and fix things" is the new medical imperative.

Journal of experimental orthopaedics·2026
Same author

Anatomically guided latent diffusion for high-resolution 3D chest CT synthesis.

Scientific reports·2026
Same author

DGPDL: Domain-Guided Prompt Distribution Learning for Generalizable Face Anti-spoofing.

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

Physically Informed 3D Food Reconstruction: Methods and Results.

IEEE journal of biomedical and health informatics·2026
Same author

VolE: A point-cloud framework for food 3D reconstruction and volume estimation.

Scientific reports·2026
Same author

Preservation of preoperative CPAK is not associated with improved clinical outcomes one-year after functionally aligned robot-assisted TKA.

Journal of orthopaedics·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 novel Error-Correcting Output Codes (ECOC) strategy for multi-class classification. By splitting classes into sub-classes, it improves performance on complex problems with overlapping data.

Area of Science:

  • Machine Learning
  • Computer Science

Background:

  • Error-Correcting Output Codes (ECOC) are a common method for multi-class classification.
  • ECOC assigns a unique codeword to each class, with classification based on codeword similarity.
  • Existing ECOC methods struggle with complex decision boundaries and class overlaps, as base classifiers may fail to separate subgroups effectively.

Purpose of the Study:

  • To propose a novel ECOC strategy utilizing sub-class information for enhanced multi-class classification.
  • To address limitations of current ECOC methods in handling non-linear decision surfaces and class overlaps.
  • To improve classification performance on complex datasets.

Main Methods:

  • A new strategy for multi-class classification using sub-class information within the ECOC framework.

Related Experiment Videos

  • Splitting original classes into sub-classes to create a problem-dependent ECOC design.
  • Embedding binary problems within this new ECOC structure.

    • Main Results:

    • The proposed sub-class splitting procedure demonstrates improved performance in ECOC.
    • Enhanced accuracy is observed when class overlap or data distribution obscures decision boundaries for the base classifier.
    • Performance gains are particularly significant with larger training dataset sizes.

    Conclusions:

    • The novel ECOC strategy effectively models complex multi-class classification problems.
    • Utilizing sub-class information and problem-dependent ECOC design enhances classifier robustness.
    • This approach offers a significant improvement over traditional ECOC methods, especially for challenging datasets.