• Journal of the Korean Data and Information Science Society
• /
• 제27권4호
• /
• pp.911-923
• /
• 2016

빅데이터 시대에 머신러닝의 중요성은 더욱 부각되고 있고 로지스틱 회귀는 머신러닝에서 분류를 위한 방법으로 의료, 경제학, 마케팅 및 사회과학 전반에 걸쳐 널리 사용되고 있다. 지금까지 R과 Hadoop의 통합환경인 RHIPE 플랫폼은 설치 및 MapReduce 구현의 어려움으로 인해 거의 연구가 이루지 지지 않았다. 본 논문에서는 대용량 데이터에 대해 로지스틱 회귀 추정을 위한 두가지 알고리즘 즉, Gradient Descent 알고리즘과 Newton-Raphson 알고리즘에 대해 MapReduce로 구현하고, 실제 데이터와 모의실험 데이터를 가지고 이들 알고리즘 간의 성능을 비교하고자 한다. 알고리즘 성능 실험에서 Gradient Descent 알고리즘은 학습률에 크게 의존하고 또한 데이터에 따라 수렴하지 않는 문제를 갖고 있다. Newton-Raphson 알고리즘은 학습률이 불필요 할 뿐만 아니라 모든 실험 데이터에 대해 좋은 성능을 보였다.

• 제어로봇시스템학회:학술대회논문집
• /
• 제어로봇시스템학회 2003년도 ICCAS
• /
• pp.134-137
• /
• 2003

Newton-Raphson is most used algorithm in EIT(electrical impedance tomography), cross-sectional distribution of resistivity is reconstructed by mean of both generating and sensing electrodes attached onto the surface of the object. EIT has been suffered from the severe ill-posedness which is caused by the inherent low sensitivity of boundary measurements to any changes of internal resistivity values. In this paper, we propose modified cost function and weighting factor that compensate for low sensitivity between boundary measurements and internal resistivity and improve performance of Newton-Raphson for EIT in region of interest.

• 한국자동차공학회논문집
• /
• 제9권1호
• /
• pp.156-162
• /
• 2001

An effective and practical method is presented for solving the forward kinematics of a 6-DOF Stewart Platform, using neural network algorithm together with Newton-Raphson method. An approximated solution is obtained from trained neural network, then it is used as an initial estimate for Newton-Raphson method. A series of accurate solutions are calculated with reasonable speed for the entire workspace of the platform. The solution procedure can be used for driving a real-time simulation platform.

• Computers and Concrete
• /
• 제10권2호
• /
• pp.121-133
• /
• 2012

Structures suffer from damages in their lifetime due to time-dependant effects, such as fatigue, creep and shrinkage, which can be expressed by concrete strains. These processes could be seen in the context of strain estimation of pre-stressed structures in two phases by using numerical methods. Their aim is checking and validating existing code procedures in determination of deformations of pre-tensioned girders by solving a system of nonlinear equations with strains as unknown parameters. This paper presents an approach based on the Newton-Raphson method for obtaining the stresses and strains in middle span section of pre-stressed girders according the equilibrium state.

• Structural Engineering and Mechanics
• /
• 제21권2호
• /
• pp.121-136
• /
• 2005

This paper presents the implementation of the Blatz-Ko hyperelastic compressible model in a finite element program to deal with large deformation problems. We show analytically and numerically that the minimum number of increment steps in the Newton-Raphson algorithm depends on material properties and applied loads. We also show that this dependence is related to the orientation preservation principle. So we propose a convergence criteria based on the sign of eigenvalues of the deformation gradient matrix.

• 산업기술연구
• /
• 제15권
• /
• pp.203-208
• /
• 1995

This paper suggests the three dimensional steady state searching techniques of hard disk/head system, which has some skew angle with flexure. In order to analyze the steady state behaviors of magnetic head slider, the localized Knudsen number and the localized bearing numbers are sued. For finding the steady state of magnetic head slider under the pre-loaded condition, I proposed multi-dimensional Newton-Raphson method which traces the equilibrium position of magnetic head slider, which has 3-degrees of freedom, using Jacobian matrix. The multi-dimensional Newton-Raphson method is very efficient technique for finding the steady state position of magnetic head slider because it approaches to the equilibrium position with changing three parameters simultaneously.

• 한국공간구조학회:학술대회논문집
• /
• 한국공간구조학회 2006년도 춘계 학술발표회 논문집 제3권1호(통권3호)
• /
• pp.170-174
• /
• 2006

This paper presents a new nonlinear analysis algorithm which uses the equivalent nodal load for the element stiffness. The equivalent nodal load represents the influence of the stiffness change such as the addition of elements, the deletion of elements, and/or the partial change of element stiffness. The nonlinear analysis of structures using the equivalent load improves the efficiency very much because the inverse of the structural stiffness matrix, which needs a large amount of computation to calculate, is reused in each loading step. In this paper, the concept of nonlinear analysis using the equivalent load for the element stiffness is described and some numerical examples are provided to verify it.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 1998년도 추계학술대회 논문집 학회본부A
• /
• pp.259-262
• /
• 1998

This paper presents a new eigenvalue calculation methods in AESOPS algorithm. The source program of the AESOPS algorithm is modified to practice in PC environment. Window95 is used as an operating system of PC and MicroSoft Power Station is used to compile the fortran source program. The heuristically approximated eigenvalue calculation method of the AESOPS algorithm is transformed to the Newton Raphson Method which is largely used in the nonlinear numerical analysis. The new methods are developed from the AESOPS algorithm and thus only a few calculation steps are added to practice the proposed algorithm.

• 한국정보과학회:학술대회논문집
• /
• 한국정보과학회 2007년도 가을 학술발표논문집 Vol.34 No.2 (B)
• /
• pp.198-201
• /
• 2007

본 논문에서는 변형된 Newton-Raphson 알고리즘과 LUT(Look Up Table)를 사용하는 역제곱근 연산기를 제안한다. Newton-Raphson 부동소수점 역수 알고리즘은 일정한 횟수의 곱셈을 반복하여 역수 제곱근을 계산하는 방식이다. 변형된 Newton-Raphson 알고리즘은 하드웨어 구현에 적합하도록 변환되었으며, LUT는 오차를 줄이기 위해 개선되었다. 제안된 연산기는 LUT의 크기를 최소화하고, 순환적인 구조가 아닌 2-stage pipeline 구조를 가진다. 또한 IEEE-754 부동소수점 표준을 기초로 하는 24-bit 데이터 형식을 사용해 면적과 속도 향상에 유리하여 휴대용 기기의 멀티미디어 분야의 응용에 적합하다. 본 역제곱근 연산기는 소수점 이하 8-bit의 정확도를 가지며 VHDL을 이용하여 설계되었다. 그 크기는 $0.18{\mu}m$ CMOS 공정에서 약 4,000 gate의 크기를 보였으며 150MHz에서 동작이 가능하다.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 2000년도 하계학술대회 논문집 A
• /
• pp.274-275
• /
• 2000

This paper introduces new ordering algorithms using the graph of data structure and forward/backward substitution of LU decomposition using recursive function. The performance of the algorithm is compared with Tinney's algorithm using 14 bus systems. Test results show that the new fill-in element of Jacobian matrix using the proposed ordering algorithm is same as that of Tinner scheme 3 and the forward/backward substitution can reduce the computation time