Search research articles

ABOUT JoVE

Overview Leadership Blog JoVE Help Center

AUTHORS

Publishing Process Editorial Board Scope & Policies Peer Review FAQ Submit

LIBRARIANS

Testimonials Subscriptions Access Resources Library Advisory Board FAQ

RESEARCH

JoVE Journal Methods Collections JoVE Encyclopedia of Experiments Archive

EDUCATION

JoVE Core JoVE Business JoVE Science Education JoVE Lab Manual Faculty Resource Center Faculty Site

Terms & Conditions of Use

Search research articles

Related Experiment Videos

Improvements to the SMO algorithm for SVM regression.

S K Shevade¹, S S Keerthi, C Bhattacharyya

¹Department of Computer Science and Automation, Indian Institute of Science, Bangalore 560012, India. shirish@csa.iisc.ernet.in

IEEE Transactions on Neural Networks

|February 6, 2008

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

Plasma creatine phosphokinase in schizophrenia.

Indian journal of psychiatry·2011

Same author

Undergraduate medical education in psychiatry - are we meeting the demands?

Indian journal of psychiatry·2011

Same author

A fast iterative nearest point algorithm for support vector machine classifier design.

IEEE transactions on neural networks·2008

Same author

Efficient tuning of SVM hyperparameters using radius/margin bound and iterative algorithms.

IEEE transactions on neural networks·2008

Same author

A stochastic connectionist approach for global optimization with application to pattern clustering.

IEEE transactions on systems, man, and cybernetics. Part B, Cybernetics : a publication of the IEEE Systems, Man, and Cybernetics Society·2008

Same author

Parallel sequential minimal optimization for the training of support vector machines.

IEEE transactions on neural networks·2006

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

This study enhances support vector machine (SVM) regression by modifying the sequential minimal optimization (SMO) algorithm. Using two threshold parameters instead of one significantly speeds up regression analysis.

Area of Science:

Machine Learning
Computational Statistics

Background:

The sequential minimal optimization (SMO) algorithm is widely used for training support vector machines (SVMs).
A known inefficiency exists in the SMO algorithm when applied to SVM regression due to a single threshold parameter.

Purpose of the Study:

To identify and address the source of inefficiency in the SMO algorithm for SVM regression.
To develop modified SMO algorithms for regression that improve computational performance.

Main Methods:

The study analyzes the Karush-Kuhn-Tucker (KKT) conditions for the dual problem of SVM regression.
Two threshold parameters are introduced to modify the standard SMO algorithm.

Main Results:

The modified SMO algorithms incorporate two threshold parameters, directly addressing the identified inefficiency.

Related Experiment Videos

Empirical results demonstrate that the modified algorithms achieve significantly faster performance compared to the original SMO on tested datasets.

Conclusions:

Modifying the SMO algorithm for SVM regression by employing two threshold parameters effectively resolves an important source of inefficiency.
The proposed modifications offer a substantial speed improvement for SVM regression tasks.