A new maximal-margin spherical-structured multi-class support vector machine

Hao, Pei-Yi; Chiang, Jung-Hsien; Lin, Yen-Hsiu

doi:10.1007/s10489-007-0101-z

Cited by 52 publications

(19 citation statements)

References 14 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…As per previous studies [22][23][24][25][26][27][28][29][30], the bridging of the MEB method with other algorithms is advantageous in the following ways: a simple algorithm is generated, time consumed is short, and robustness is enhanced. However, the datasets of realworld classification problems generally exhibit distinctive distributions, and the ball models cannot describe these problems ideally.…”

Section: Introductionmentioning

confidence: 94%

“…Through the data description of MEB computation in the training datasets, the authors obtained approximately optimal solutions for large datasets. Hao et al [23] established a maximal-margin, spherical-structured, multi-class SVM by incorporating the concept of maximal margin into spherical structures. A geometric insight query method was proposed in [24] based on MEB, convex hull, and the furthest Voronoi diagram of the query group for Group enclosing query (GEQ) problems.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

A new multi-class classification method based on minimum enclosing balls

et al. 2015

View full text Add to dashboard Cite

With respect to classification problems, the Minimum enclosing ball (MEB) method was recently studied by some scholars as a new support vector machine. As a nascent technology, however, MEB reports poor adaptability for different types of samples, especially multi-class samples. In this paper, we propose a new multi-class classification method based on MEB. This method is derived from each class sample center and radius with the Gaussian kernel width factor parameter σ, which is labelled as σ-MEB. σ is a variable parameter according to the different sample characteristics. When this parameter is considered, the multi-class classifier is easy to adapt and is robust in diverse datasets. The quadratic programming problem was transformed into its dual form with Lagrange multipliers using this method. Finally, we applied sequential minimal optimization method and Karush-Kuhn-Tucker conditions to accelerate the training process. Numerical experiment results indicate that for given different types of samples, the proposed method is more accurate than the methods with which it is compared. Moreover, the proposed method reports values in the upper quantile with respect to adaptation performance.

show abstract

Section: Introductionmentioning

confidence: 94%

Section: Introductionmentioning

confidence: 99%

A new multi-class classification method based on minimum enclosing balls

et al. 2015

View full text Add to dashboard Cite

show abstract

“…In the training stage, training samples after normalization and scale processing (Hao, et al 2007) are fed to a SVM with indicating their corresponding class. The features are computed from the training data, each contains vector from the training fingerprint, and the identity ID of the corresponding class is used to guide the classifying results through the SVM.…”

Section: Performances With a Svm Matchingmentioning

confidence: 99%

Non-minutiae based fingerprint descriptor

Yang¹

2011

Biometrics

View full text Add to dashboard Cite

“…There are many works in the literature that exploit spherical, ellipsoidal, and radial basis surfaces for classification purposes [16][17][18][19][20][21][22][23][24][25][26]. However, some of them are devoted to a single-class classification problem only.…”

Section: Introductionmentioning

confidence: 99%

A penalty function method for designing efficient robust classifiers with input space optimal separating surfaces

Uçar¹,

Demir²,

Güzeliş³

2014

Turk J Elec Eng & Comp Sci

View full text Add to dashboard Cite

This paper considers robust classification as a constrained optimization problem. Where the constraints are nonlinear, inequalities defining separating surfaces, whose half spaces include or exclude the data depending on their classes and the cost, are used for attaining robustness and providing the minimum volume regions specified by the half spaces of the surfaces. The constraints are added to the cost using penalty functions to get an unconstrained problem for which the gradient descent method can be used. The separating surfaces, which are aimed to be found in this way, are optimal in the input data space in contrast to the conventional support vector machine (SVM) classifiers designed by the Lagrange multiplier method, which are optimal in the (transformed) feature space. Two types of surfaces, namely hyperellipsoidal and Gaussian-based surfaces created by radial basis functions (RBFs), are focused on in this paper due to their generality. Ellipsoidal classifiers are trained in 2 stages: a spherical surface is found in the first stage, and then the centers and the radii found in the first stage are taken as the initial input for the second stage to find the center and covariance matrix parameters of the ellipsoids. The penalty function approach to the design of robust classifiers enables the handling of multiclass classification. Compared to SVMs, multiple-kernel SVMs, and RBF classifiers, the proposed classifiers are found to be more efficient in terms of the required training time, parameter setting time, testing time, memory usage, and generalization error, especially for medium to large datasets. RBF-based input space optimal classifiers are also introduced for problems that are far from ellipsoidal, e.g., 2 Spirals.

show abstract

A new maximal-margin spherical-structured multi-class support vector machine

Cited by 52 publications

References 14 publications

A new multi-class classification method based on minimum enclosing balls

A new multi-class classification method based on minimum enclosing balls

Non-minutiae based fingerprint descriptor

A penalty function method for designing efficient robust classifiers with input space optimal separating surfaces

Contact Info

Product

Resources

About