Search research articles

Search research articles

Related Experiment Videos

Fuzzy least squares support vector machines for multiclass problems.

Daisuke Tsujinishi¹, Shigeo Abe

¹Graduate School of Science and Technology, Kobe University, Rokkodai, Nada, Kobe, Japan. tujinisi@frenchblue.scitec.kobe-u.ac.jp

Neural Networks : the Official Journal of the International Neural Network Society

|July 10, 2003

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

Impact of routine recurrent laryngeal nerve monitoring in prone esophagectomy with mediastinal lymph node dissection.

Surgical endoscopy·2016

Same author

Murine double minute 2 predicts response of advanced esophageal squamous cell carcinoma to definitive chemoradiotherapy.

BMC cancer·2015

Same author

Comparison of short-term outcomes between prone and lateral decubitus positions for thoracoscopic esophagectomy.

Surgical endoscopy·2014

Same author

Complications following recurrent laryngeal nerve lymph node dissection in oesophageal cancer surgery.

Interactive cardiovascular and thoracic surgery·2014

Same author

A strategy for supraclavicular lymph node dissection using recurrent laryngeal nerve lymph node status in thoracic esophageal squamous cell carcinoma.

The Annals of thoracic surgery·2013

Same author

Murine double minute 2 and its association with chemoradioresistance of esophageal squamous cell carcinoma.

Anticancer research·2013

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

Fuzzy least squares support vector machines (LS-SVMs) address multiclass problems by defining membership functions. Fuzzy LS-SVMs with a minimum operator show performance comparable to fuzzy SVMs.

Area of Science:

Machine Learning
Artificial Intelligence
Pattern Recognition

Background:

Least Squares Support Vector Machines (LS-SVMs) offer an efficient alternative to traditional SVMs by solving linear equations.
Standard SVMs and LS-SVMs are inherently binary classifiers, leading to challenges in multiclass applications.
Unclassifiable regions arise when extending binary classification methods to handle multiple classes.

Purpose of the Study:

To introduce and evaluate Fuzzy LS-SVMs for resolving unclassifiable regions in multiclass classification problems.
To adapt the LS-SVM framework for multiclass scenarios using fuzzy membership functions.
To compare the performance of different fuzzy operators within the Fuzzy LS-SVM model.

Main Methods:

Developed Fuzzy LS-SVMs by defining membership functions perpendicular to separating hyperplanes for pairwise class separation.

Related Experiment Videos

Applied minimum and average operators to combine pairwise membership functions for defining class-specific membership.

Evaluated the proposed Fuzzy LS-SVM approach on benchmark datasets.

Main Results:

Fuzzy LS-SVMs utilizing the minimum operator achieved recognition performance comparable to existing fuzzy SVMs.
Fuzzy LS-SVMs employing the average operator demonstrated inferior performance compared to the minimum operator.
The study highlights the effectiveness of the minimum operator in fuzzy LS-SVM for multiclass classification.

Conclusions:

Fuzzy LS-SVMs provide a viable solution for multiclass classification by effectively handling unclassifiable regions.
The choice of fuzzy operator significantly impacts the performance of Fuzzy LS-SVMs, with the minimum operator being more effective.
This research contributes to advancing multiclass classification techniques within the LS-SVM framework.