• Journal of the Korean Data and Information Science Society
• /
• 제25권4호
• /
• pp.951-959
• /
• 2014

Recently Hazelton and Turlach (2009) proposed a weighted kernel density estimator for the deconvolution problem. In the case of Gaussian kernels and measurement error, they argued that the weighted kernel density estimator is a competitive estimator over the classical deconvolution kernel estimator. In this paper we consider weighted kernel density estimators when sample observations are contaminated by double exponentially distributed errors. The performance of the weighted kernel density estimators is compared over the classical deconvolution kernel estimator and the kernel density estimator based on the support vector regression method by means of a simulation study. The weighted density estimator with the Gaussian kernel shows numerical instability in practical implementation of optimization function. However the weighted density estimates with the double exponential kernel has very similar patterns to the classical kernel density estimates in the simulations, but the shape is less satisfactory than the classical kernel density estimator with the Gaussian kernel.

• 대한음성학회지:말소리
• /
• 제66호
• /
• pp.105-116
• /
• 2008

In this research, we propose a speaker identification system using a kernel method which is expected to model the non-linearity of speech features well. We have been using principal component analysis (PCA) successfully, and extended to kernel PCA, which is used for many pattern recognition tasks such as face recognition. However, we cannot use kernel PCA for speaker identification directly because the storage required for the kernel matrix grows quadratically, and the computational cost grows linearly (computing eigenvector of $l{\times}l$ matrix) with the number of training vectors I. Therefore, we use greedy kernel PCA which can approximate kernel PCA with small representation error. In the experiments, we compare the accuracy of the greedy kernel PCA with the baseline Gaussian mixture models using MFCCs and PCA. As the results with limited enrollment data show, the greedy kernel PCA outperforms conventional methods.

• 대한수학회지
• /
• 제33권2호
• /
• pp.283-290
• /
• 1996

We can expect a close relationship between the Bergman kernel and the Szego kernel of a domain because we can change boundary integrals to solid integrals via Green's identity.

• 한국정보과학회:학술대회논문집
• /
• 한국정보과학회 2006년도 한국컴퓨터종합학술대회 논문집 Vol.33 No.1 (A)
• /
• pp.325-327
• /
• 2006

현재 사용되고 있는 운영체제들은 그들의 기능을 확장하거나 교체하기 위해서 kernel extension을 사용해 왔다. 일반적으로 이러한 kernel extension들은 커널과 같은 주소공간에서 실행하기 때문에, 그것에서 발생하는 오류(fault)로 인해 전체 시스템이 망가지는 결과를 초래할 위험이 있다. 그래서 kernel extension의 안전한 실행에 관한 연구들은 kernel extension에서 발생한 오류를 전체 시스템으로부터 고립시키는 것이 주목적이었다. 하지만 이러한 방법들은 kernel extension의 어셈블리어로 된 코드를 분석하거나 사용하고 있는 커널의 소스 코드를 수정을 필요로 한다. 본 논문은 Sentry라는 kernel extension을 감시하기 위한 인터포지션 서비스를 제안한다. Sentry를 사용하기 위해서 별도의 커널 코드를 수정할 필요도 없으며, 이미 사용하고 있는 리눅스와 호환될 수 있는 특징을 지니고 있다. 그리고 kernel extension의 소스코드 및 어셈블리 코드에 대한 분석 없이 바이너리 파일을 직접 수정하여 kernel extension을 모니터링 할 수 있도록 한다. 게다가 Sentry는 재구성이 가능하기 때문에 얼마든지 kernel extension에 대한 보호정책을 동적으로 바꿀 수 있다.

• 한국산학기술학회논문지
• /
• 제12권7호
• /
• pp.3251-3257
• /
• 2011

본 논문에서는 논리합성을 위한 공통식 추출 방법을 새롭게 제안한다. 제안하는 방법은 주어진 각 논리식들에서 커널/커널 쌍들과 코커널/커널 쌍을 추출한다. 커널/커널 쌍은 주어진 논리식을 부울 나눗셈에 의해 제수, 몫, 나머지로 논리식을 다시 표현하게 된다. 다음, 여러 논리식에서 산출된 제수, 몫들에서 공통식을 추출하는 커널 교집합에 의해 공통식을 구하는 방법을 제안한다. 실험 결과 기존의 공통식 산출 결과들과 비교했을 때 제안한 방법은 리터럴 개수를 줄일 수 있었다.

• 대한수학회지
• /
• 제43권1호
• /
• pp.199-213
• /
• 2006

We compute the holomorphic derivative of the harmonic measure associated to a $C^\infty$bounded domain in the plane and show that the exact Bergman kernel function associated to a $C^\infty$ bounded domain in the plane relates the derivatives of the Ahlfors map and the Szego kernel in an explicit way. We find several formulas for the exact Bergman kernel and the Szego kernel and the harmonic measure. Finally we survey some other properties of the holomorphic derivative of the harmonic measure.

• 호남수학학술지
• /
• 제27권2호
• /
• pp.195-203
• /
• 2005

We show that the exact Bergman kernel function associated to a $C^{\infty}$ bounded domain in the plane relates the derivatives of the Ahlfors map in an explicit way. And we find several formulas relating the exact Bergman kernel to classical kernel functions in potential theory.

• 한국산업융합학회 논문집
• /
• 제3권1호
• /
• pp.17-27
• /
• 2000

A factorization is an extremely important part of multi-level logic synthesis. The number of literals in a factored form is a good estimate of the complexity of a logic function. and can be translated directly into the number of transistors required for implementation. Factored forms are described as either algebraic or Boolean, according to the trade-off between run-time and optimization. A Boolean factored form contains fewer number of literals than an algebraic factored form. In this paper, we present a new method for a Boolean factorization. The key idea is to build an extended co-kernel cube matrix using co-kernel/kernel pairs and kernel/kernel pairs together. The extended co-kernel cube matrix makes it possible to yield a Boolean factored form. We also propose a heuristic method for covering of the extended co-kernel cube matrix. Experimental results on various benchmark circuits show the improvements in literal counts over the algebraic factorization based on Brayton's co-kernel cube matrix.

• 한국CDE학회논문집
• /
• 제6권4호
• /
• pp.271-276
• /
• 2001

A geometric kernel is the library of core mathematical functions that defines and stores 3D shapes in response to users'commands. We developed a light geometric kernel suitable to develop CAD/CAM application systems. The kernel contains geometric objects, such as points, curves and surfaces and a minimal set of functions for each type but does not contain lots of modeling and handling functions that are useful to create and maintain complex shapes from an idea sketch. The kernel was developed on MS-Windows NT using C++ with STL(Standard Template Library) but it is compatible with UNIX environments. This paper describes the structure of the kernel including several components: base, math, point sequence curve, geometry, translators. The base kernel gives portability to applications and the math kernel contains basic arithmetic and their classes, such as vector and matrix. The geometry kernel contains points, parametric curves, and parametric surfaces. A neutral fie format and programming and document styles are also presented in this paper.

• 대한임베디드공학회논문지
• /
• 제13권4호
• /
• pp.187-194
• /
• 2018

The operating system for IoT should have a small memory footprint and provide low power state, real-time, multitasking, various network protocols, and security. Although the Zephyr kernel, an operating system for IoT, released by the Linux Foundation in February 2016, has these features but errors generated by the user code can generate fatal problems in the system because the Zephyr kernel adopts a single-space method that both the user code and kernel code execute in the same space. In this research, we propose a space separation method, which separates kernel space and user space, to solve this problem. The space separation that we propose consists of three modifications in Zephyr kernel. The first is the code separation that kernel code and user code execute in each space while using different stacks. The second is the kernel space protection that generates an exception by using the MPU (Memory Protection Unit) when the user code accesses the kernel space. The third is the SVC based system call that executes the system call using the SVC instruction that generates the exception. In this research, we implemented the space separation in Zephyr v1.8.0 and evaluated safety through abnormal execution of the user code. As the result, the kernel was not crashed by the errors generated by the user code and was normally executed.