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

Training hard-margin support vector machines using greedy stagewise algorithm.

Liefeng Bo1, Ling Wang, Licheng Jiao

  • 1Key Laboratory of Intelligent Perception and Image Understanding of Ministry of Education of China, the Institute of Intelligent Information Processing, Xidian University, Xi'an, Shaanxi 710071, P. R. China. blf0218@ 163.com

IEEE Transactions on Neural Networks
|August 15, 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

Wavelet spectral-aware Kolmogorov-Arnold Network for organ and tumor segmentation.

Computerized medical imaging and graphics : the official journal of the Computerized Medical Imaging Society·2026
Same author

Data and knowledge-driven imaging biomarkers for lumbar aging and degenerative risk stratification monitoring.

NPJ digital medicine·2026
Same author

Scale-Aware Prompting With Optimal Transport for Remote Sensing Image Captioning.

IEEE transactions on image processing : a publication of the IEEE Signal Processing Society·2026
Same author

Learning Evolution Via Optimization Knowledge Adaptation.

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

DI3CL: Contrastive Learning With Dynamic Instances and Contour Consistency for SAR Land-Cover Classification Foundation Model.

IEEE transactions on image processing : a publication of the IEEE Signal Processing Society·2026
Same author

Coarse-to-Fine Fusion: Customized Multiview Contrast Reinforcement Learning for Graph Clustering.

IEEE transactions on neural networks and learning systems·2026
Same journal

Universal perceptron and DNA-like learning algorithm for binary neural networks: LSBF and PBF implementations.

IEEE transactions on neural networks·2013
Same journal

Guest editorial: special section on white box nonlinear prediction models.

IEEE transactions on neural networks·2011
Same journal

Data-based fault-tolerant control of high-speed trains with traction/braking notch nonlinearities and actuator failures.

IEEE transactions on neural networks·2011
Same journal

Guest editorial: special section on data-based control, modeling, and optimization.

IEEE transactions on neural networks·2011
Same journal

Neural network-based multiple robot simultaneous localization and mapping.

IEEE transactions on neural networks·2011
Same journal

Data-driven model-free adaptive control for a class of MIMO nonlinear discrete-time systems.

IEEE transactions on neural networks·2011
See all related articles

A new greedy stagewise algorithm for Support Vector Machines (GS-SVMs) effectively combats overfitting without regularization. This method offers a computationally efficient alternative for large datasets, matching accuracy with reduced processing time.

Area of Science:

  • Machine Learning
  • Computational Statistics

Background:

  • Hard-margin Support Vector Machines (HM-SVMs) are prone to overfitting with noisy data.
  • Soft-margin SVMs mitigate overfitting using regularization but incur high computational costs.

Purpose of the Study:

  • Introduce a novel greedy stagewise algorithm for SVMs (GS-SVMs) to address HM-SVM overfitting without regularization.
  • Evaluate the computational efficiency and accuracy of GS-SVMs on large datasets.

Main Methods:

  • Developed a greedy stagewise algorithm for SVMs (GS-SVMs).
  • Utilized statistical learning theory to analyze the algorithm's performance.
  • Conducted experiments on datasets with up to 400,000 training samples.

Main Results:

Related Experiment Videos

  • GS-SVMs effectively handle overfitting without explicit regularization terms.
  • The computational complexity of GS-SVMs scales quadratically with training samples.
  • Experiments show GS-SVMs are faster than LIBSVM 2.83 on large datasets without accuracy loss.

Conclusions:

  • GS-SVMs provide an efficient and accurate method for training SVMs on large, potentially noisy datasets.
  • The early stopping rule in GS-SVMs functions as an implicit regularization mechanism.
  • This approach offers a viable alternative to traditional soft-margin SVMs.