• Proceedings of the Korean Information Science Society Conference
• /
• 2004.10b
• /
• pp.703-705
• /
• 2004

Conventional support vector machines (SVMs) find optimal hyperplanes that have maximal margins by treating all data equivalently. In the real world, however, the data within a data set may differ in degree of uncertainty or importance due to noise, inaccuracies or missing values in the data. Hence, if all data are treated as equivalent, without considering such differences, the optimal hyperplanes identified are likely to be less optimal. In this paper, to more accurately identify the optimal hyperplane in a given uncertain data set, we propose a membership-induced distance from a hyperplane using membership values, and formulate three kinds of membership-induced SVMs.

• Proceedings of the Korean Institute of Intelligent Systems Conference
• /
• 2001.12a
• /
• pp.60-64
• /
• 2001

In this paper, a new learning methodology for Kernel Methods is suggested that results in a sparse representation of kernel space from the training patterns for classification problems. Among the traditional algorithms of linear discriminant function(perceptron, relaxation, LMS(least mean squared), pseudoinverse), this paper shows that the relaxation procedure can obtain the maximum margin separating hyperplane of linearly separable pattern classification problem as SVM(Support Vector Machine) classifier does. The original relaxation method gives only the necessary condition of SV patterns. We suggest the sufficient condition to identify the SV patterns in the learning epochs. Experiment results show the new methods have the higher or equivalent performance compared to the conventional approach.

• Bulletin of the Korean Mathematical Society
• /
• v.26 no.2
• /
• pp.185-190
• /
• 1989

In [4], J. Leray introduced the notion of partial hyperbolicity to characterize the operators for which the non-characteristic Cauchy problem is solvable in the Geverey class for any data which are holomorphic in a part of variables x"=(x$_{2}$,..,x$_{l}$ ) in the initial hyperplane x$_{1}$=0. A linear partial differential operator is called partially hyperbolic modulo the linear subvarieties S:x"=constant if the equation P$_{m}$(x, .zeta.$_{1}$, .xi.')=0 for .zeta.$_{1}$ has only real roots when .xi.'is real and .xi."=0, where P$_{m}$ is the principal symbol of pp. Limiting to the case of operators with constant coefficients, A. Kaneko proposed a new sharper condition when S is a hyperplane [3]. In this paper, we generalize this condition to the case of general linear subvariety S and show that it is sufficient for the solvability of Cauchy problem for the hyperfunction Cauchy data which contains variables parallel to S as holomorphic parameters.blem for the hyperfunction Cauchy data which contains variables parallel to S as holomorphic parameters.

• Bulletin of the Korean Mathematical Society
• /
• v.50 no.3
• /
• pp.823-832
• /
• 2013

We present a new algorithm for solving a system of nonlinear equations with convex constraints which combines proximal point and projection methodologies. Compared with the existing projection methods for solving the problem, we use a different system of linear equations to obtain the proximal point; and moreover, at the step of getting next iterate, our projection way and projection region are also different. Based on the Armijo-type line search procedure, a new hyperplane is introduced. Using the separate property of hyperplane, the new algorithm is proved to be globally convergent under much weaker assumptions than monotone or more generally pseudomonotone. We study the convergence rate of the iterative sequence under very mild error bound conditions.

• The Korean Journal of Applied Statistics
• /
• v.17 no.1
• /
• pp.75-88
• /
• 2004

The Support Vector Machine(SVM) algorithm has paid attention on maximizing the shortest distance between sample points and discrimination hyperplane. This paper suggests the total margin algorithm which considers the distance between all data points and the separating hyperplane. The method extends existing support vector machine algorithm. In addition, this newly proposed method improves the generalization error bound. Numerical experiments show that the total margin algorithm provides good performance, comparing with the previous methods.

• Proceedings of the KSME Conference
• /
• 2000.04a
• /
• pp.469-474
• /
• 2000

A sliding mode control method using the parameterization of both the hyperplane and the compensator for output feedback and reduced observer is presented for rotational inverted pendulums. This control strategy overcomes the problem of unattainable velocity state which is resulted from severe noise of analogue sense and constructs numerical algorithms designs of dynamic output feedback sliding mode hyperplane and controller. The result of the experiment shows the superior performance compared with the LQ controller and the robustness with respect to both tapping disturbances and certain initial conditions.

• Bulletin of the Korean Mathematical Society
• /
• v.30 no.1
• /
• pp.29-34
• /
• 1993

J. Leray [7] proposed a sufficient condition ofr the solvability of the Cauchy problem on the initial hyperplane x$_{1}$=0 with Cauchy data which are holomorphic with respect to the variables parallel to some analytic subvariety S of the initial hyperplane. Limiting the problem to the case of operators with constant coefficients, A. Kaneko [2] proposed a new sharper sufficient condition. Later we generalized this condition and showed that it is necessary and sufficient for the solvability of the Cauchy problem for the hyperfunction Cauchy data and the distribution Cauchy data which contain variables parallel to S as holomorphic parameters in [5, 6]. In this paper, we extend the results in [6] to the case of operators with variable coefficients and show that it is sufficient for the solvability of the Cauchy problem for the hyperfunction Cauchy data. Our main theorem can be considered as an example of a deep theorem on micro-hyperbolic systems by Kashiwara-Schapira [4] and we give a direct proof based on an elementary sweeping out procedure developed in Kaneko [3].

• Journal of the Korean Mathematical Society
• /
• v.58 no.6
• /
• pp.1485-1500
• /
• 2021

A curve is rectifying if it lies on a moving hyperplane orthogonal to its curvature vector. In this work, we extend the main result of [Chen 2017, Tamkang J. Math. 48, 209] to any space dimension: we prove that rectifying curves are geodesics on hypercones. We later use this association to characterize rectifying curves that are also slant helices in three-dimensional space as geodesics of circular cones. In addition, we consider curves that lie on a moving hyperplane normal to (i) one of the normal vector fields of the Frenet frame and to (ii) a rotation minimizing vector field along the curve. The former class is characterized in terms of the constancy of a certain vector field normal to the curve, while the latter contains spherical and plane curves. Finally, we establish a formal mapping between rectifying curves in an (m + 2)-dimensional space and spherical curves in an (m + 1)-dimensional space.

• Journal of IKEEE
• /
• v.13 no.2
• /
• pp.140-149
• /
• 2009

Traffic anomaly detection is one of important technology that should be considered in network security and administration. In this paper, we propose an abnormal traffic detection mechanism that includes traffic monitoring and traffic analysis. We develop analytical passive monitoring system called WISE-Mon which can inspect traffic behavior. We establish a criterion by analyzing the characteristics of a traffic training set. To detect abnormal traffic, we derive a hyperplane by using Fisher linear discriminant and chi-square distribution as well as the analyzed characteristics of traffic. Our mechanism can support reliable results for traffic anomaly detection and is compatible to real-time detection. In addition, since the trend of traffic can be changed as time passes, the hyperplane has to be updated periodically to reflect the changes. Accordingly, we consider the self-learning algorithm which reflects the trend of the traffic and so enables to increase the pliability of detection probability. Numerical results are presented to validate the accuracy of proposed mechanism. It shows that the proposed mechanism is reliable and relevant for traffic anomaly detection.

• Journal of KIISE:Software and Applications
• /
• v.27 no.4
• /
• pp.422-431
• /
• 2000

This paper proposes a new codebook generation method, called a PCA-Based VQ, that incorporates the PCA (Principal Component Analysis) technique into VQ (Vector Quantization) codebook design. The PCA technique reduces the data dimensions by transforming input image vectors into the feature vectors. The cluster of feature vectors in the transformed domain is bisectioned into two subclusters by an optimally chosen partitioning hyperplane. We expedite the searching of the optimal partitioning hyperplane that is the most time consuming process by considering that (1) the optimal partitioning hyperplane is perpendicular to the first principal axis of the feature vectors, (2) it is located on the equilibrium point of the left and right cluster's distortions, and (3) the left and right cluster's distortions can be adjusted incrementally. This principal axis bisectioning is successively performed on the cluster whose difference of distortion between before and after bisection is the maximum among the existing clusters until the total distortion of clusters becomes as small as the desired level. Simulation results show that the proposed PCA-based VQ method is promising because its reconstruction performance is as good as that of the SOFM (Self-Organizing Feature Maps) method and its codebook generation is as fast as that of the K-means method.