• 응용통계연구
• /
• 제36권5호
• /
• pp.381-397
• /
• 2023

대부분의 임상시험에서 가장 대표적으로 사용되는 실험설계는 무작위화로, 치료 효과를 정확하게 추정하기 위해 이용된다. 그러나 무작위화가 이루어지지 않은 관찰연구의 경우 치료군과 대조군의 비교로 얻는 치료효과에는 환자 간의 특성 등 여러 조정되지 않은 차이로 인해 편향이 발생한다. 성향 점수 가중치는 이러한 문제점을 해결하기 위해 널리쓰이는 방법으로 치료 효과를 추정하는데에 있어 교란요인을 조정하여 편향을 최소화하도록 하는 방법이다. 성향 점수를 이용한 가중치 방법 중 가장 널리 알려진 역확률 가중치는 관찰된 공변량이 주어졌을 때 특정 치료에 할당될 조건부 확률의 역에 비례하는 가중치를 할당한다. 그러나 이 방법은 극단적인 성향 점수에 의해 종종 방해 받아 편향된 추정치와 과도한 분산을 초래한다는 점이 알려져있어 이러한 문제를 완화하기 위해 절사 역확률 가중치, 중복 가중치, 일치 가중치를 포함한 여러 가지 대안 방법이 제안되었다. 본 논문에서는 제한된 중복, 잘못 지정된 성향 점수 모델 및 예측과 반대되는 치료 등 다양한 문제상황에서 여러 성향 점수 가중치 방법의 성능을 비교하는 시뮬레이션 비교연구를 수행하였다. 비교연구의 결과 중복 가중치와 일치 가중치는 편향, 제곱근평균제곱오차 및 포함 확률 측면에서 역확률 가중치와 절사역확률 가중치에 비에 우월한 성능을 보임을 확인하였다.

• Smart Structures and Systems
• /
• 제8권4호
• /
• pp.343-365
• /
• 2011

Structural system identification (SSI) is an inverse problem of difficult solution. Currently, difficulties lie in the development of algorithms which can cater to large size problems. In this paper, a parameter estimation technique based on evolutionary strategy is presented to overcome some of the difficulties encountered in using the traditional system identification methods in terms of convergence. In this paper, a non-traditional form of system identification technique employing evolutionary algorithms is proposed. In order to improve the convergence characteristics, it is proposed to employ immune algorithms which are proved to be built with superior diversification mechanism than the conventional evolutionary algorithms and are being used for several practical complex optimisation problems. In order to reduce the number of design variables, domain decomposition methods are used, where the identification process of the entire structure is carried out in multiple stages rather than in single step. The domain decomposition based methods also help in limiting the number of sensors to be employed during dynamic testing of the structure to be identified, as the process of system identification is carried out in multiple stages. A fifteen storey framed structure, truss bridge and 40 m tall microwave tower are considered as a numerical examples to demonstrate the effectiveness of the domain decomposition based structural system identification technique using immune algorithm.

• Journal of applied mathematics & informatics
• /
• 제26권1_2호
• /
• pp.213-222
• /
• 2008

Iterative methods are often suitable for solving least squares problems min$||Ax-b||_2$, where A $\epsilon\;\mathbb{R}^{m{\times}n}$ is large and sparse. The well known LSQR algorithm is among the iterative methods for solving these problems. A good preconditioner is often needed to speedup the LSQR convergence. In this paper we present the numerical experiments of applying a well known preconditioner for the LSQR algorithm. The preconditioner is based on the $A^T$ A-orthogonalization process which furnishes an incomplete upper-lower factorization of the inverse of the normal matrix $A^T$ A. The main advantage of this preconditioner is that we apply only one of the factors as a right preconditioner for the LSQR algorithm applied to the least squares problem min$||Ax-b||_2$. The preconditioner needs only the sparse matrix-vector product operations and significantly reduces the solution time compared to the unpreconditioned iteration. Finally, some numerical experiments on test matrices from Harwell-Boeing collection are presented to show the robustness and efficiency of this preconditioner.

• Journal of the Korean Society for Industrial and Applied Mathematics
• /
• 제5권1호
• /
• pp.85-100
• /
• 2001

In this paper we propose a block-type parallel preconditioner for solving large sparse nonsymmetric linear systems, which we expect to be scalable. It is Multi-Color Block SOR preconditioner, combined with direct sparse matrix solver. For the Laplacian matrix the SOR method is known to have a nondeteriorating rate of convergence when used with Multi-Color ordering. Since most of the time is spent on the diagonal inversion, which is done on each processor, we expect it to be a good scalable preconditioner. Finally, due to the blocking effect, it will be effective for ill-conditioned problems. We compared it with four other preconditioners, which are ILU(0)-wavefront ordering, ILU(0)-Multi-Color ordering, SPAI(SParse Approximate Inverse), and SSOR preconditioner. Experiments were conducted for the Finite Difference discretizations of two problems with various meshsizes varying up to 1024 x 1024, and for an ill-conditioned matrix from the shell problem from the Harwell-Boeing collection. CRAY-T3E with 128 nodes was used. MPI library was used for interprocess communications. The results show that Multi-Color Block SOR and ILU(0) with Multi-Color ordering give the best performances for the finite difference matrices and for the shell problem only the Multi-Color Block SOR converges.

• 한국군사과학기술학회지
• /
• 제14권2호
• /
• pp.313-320
• /
• 2011

SPH(Smoothed Particle Hydrodynamics) is a gridless Lagrangian technique that is useful as an alternative numerical analysis method used to analyze high deformation problems as well as astrophysical and cosmological problems. In SPH, all points within the support of the kernel are taken as neighbours. The accuracy of the SHP is highly influenced by the method for choosing neighbours from all particle points considered. Typically a linked-list method or tree search method has been used as an effective tool because of its conceptual simplicity, but these methods have some liability in anisotropy situations. In this study, convex hull algorithm is presented as an improved method to eliminate this artifact. A convex hull is the smallest convex set that contains a certain set of points or a polygon. The selected candidate neighbours set are mapped into the new space by an inverse square mapping, and extract a convex hull. The neighbours are selected from the shell of the convex hull. These algorithms are proved by Fortran programs. The programs are expected to use as a searching algorithm in the future SPH program.

• 대한수학교육학회지:학교수학
• /
• 제4권2호
• /
• pp.205-222
• /
• 2002

In school mathematics, the negative numbers have been instructed using the intuitive models such as the number line model, the counting model, and inductive-extrapolation on the additionand multiplication and using inverse operation on the subtraction and division. Theseinstructions on the negative numbers did not present their formal nature and caused the difficulty for students to understand their operations because of the incomplete function of the intuitive models. In this study, we tried to improve such problems of the instructions of the negative numbers on the basis of the didactical phenomenological analysis. First of all, we analysed the nature of the negative numbers and the cognitive obstructions through the examination about the historic process of them. Second, we examined hew the nature of the negative numbers were analysed and described in mathematics. Third, we explored the improving directions for them on the ground of the didactical phenomenological analysis. In school mathematics, the rules of operations using the intuitive models of the negative numbers have been Instructed rather than approaching toward the nature of them. The negative numbers have been developed from the necessity to find the general solution of equations. The study tries to approach the operations instructions of the negative numbers formative]y to overcome the problems of those that are using the intuitive models and to reflect the formative Furthermore of the negative numbers. Furthermore, we examine the way of the instruction of the negative numbers in real context so that the algebraic feature and the real context should be Interactive.

• 대한의용생체공학회:의공학회지
• /
• 제29권3호
• /
• pp.231-238
• /
• 2008

Nelder-Mead downhill simplex method (DSM), a kind of deterministic optimization algorithms, has been used extensively for magnetoencephalography(MEG) dipolar source localization problems because it dose not require any functional differentiation. Like many other deterministic algorithms, however, it is very sensitive to the choice of initial positions and it can be easily trapped in local optima when being applied to complex inverse problems with multiple simultaneous sources. In this paper, some modifications have been made to make up for DSM's limitations and improve the accuracy of DSM. First of all, initial point determination method for DSM using magnetic fields on the sensor surface was proposed. Secondly, Univariant-DSM combined DSM with univariant method was proposed. To verify the performance of the proposed method, it was applied to simulated MEG data and practical MEG measurements.

• 한국수학교육학회지시리즈A:수학교육
• /
• 제49권2호
• /
• pp.161-174
• /
• 2010

The purpose of the study is to analyze various types of errors appeared in true-false proof problems of matrix and to find out correction method. In order to achieve this purpose, error test was conducted to the subject of 87 second grade students who were chosen from D high schoool. It was shown from this test that the most frequent error type was caused by the lack of understanding about concepts and essential facts of matrix(35.3%), and then caused by the invalid logically reasoning (27.4%), and then caused by the misusing conditions(18.7%). Through three hours of correction lessons with 5 students, the following correction teaching method was proposed. First, it is stressed that the operation rules and properties satisfied in real number system can not be applied in matrix. Second, it is taught that the analytical proof method and the reductio ad absurdum method are useful in the proof problem of matrix. Third, it is explained that the counter example of E=$\begin{pmatrix}1\;0\\0\;1 \end{pmatrix}$, -E should be found in proof of the false statement. Fourth, it is taught that the determinant condition should be checked for the existence of the inverse matrix.

• 호남수학학술지
• /
• 제41권2호
• /
• pp.403-420
• /
• 2019

The Helmholtz equation represents acoustic or electromagnetic scattering phenomena. The Method of Lines are known to have many advantages in simulation of forward and inverse scattering problems due to the usage of angle rays and Bessel functions. However, the method does not account for the jump phenomena on obstacle boundary and the approximation includes many high order Bessel functions. The high order Bessel functions have extreme blow-up or die-out features in resonance region obstacle boundary. Therefore, in particular, when we consider shape reconstruction problems, the method is suffered from severe instabilities due to the logical confliction and the severe singularities of high order Bessel functions. In this paper, two approximation formulas for the Helmholtz equation are introduced. The formulas are new and powerful. The derivation is based on Method of Lines, Huygen's principle, boundary jump relations, Addition Formula, and the orthogonality of the trigonometric functions. The formulas reduce the approximation dimension significantly so that only lower order Bessel functions are required. They overcome the severe instability near the obstacle boundary and reduce the computational time significantly. The convergence is exponential. The formulas adopt the scattering jump phenomena on the boundary, and separate the boundary information from the measured scattered fields. Thus, the sensitivities of the scattered fields caused by the boundary changes can be analyzed easily. Several numerical experiments are performed. The results show the superiority of the proposed formulas in accuracy, efficiency, and stability.

• 대한수학회보
• /
• 제41권1호
• /
• pp.125-135
• /
• 2004

One of the intriguing and fundamental algorithmic graph problems is the computation of the strongly connected components of a directed graph G. In this paper we first introduce a simple procedure for determining the location of the nonzero elements occurring in $B^{-1}$ without fully inverting B, where EB\;{\equiv}\;(b_{ij)\;and\;B^T$ are diagonally dominant matrices with $b_{ii}\;>\;0$ for all i and $b_{ij}\;{\leq}\;0$, for $i\;{\neq}\;j$, and then, as an application, show that all of the strongly connected components of a directed graph G can be recognized by the location of the nonzero elements occurring in the matrix $C(G)\;=\;(D\;-\;A(G))^{-1}$. Here A(G) is an adjacency matrix of G and D is an arbitrary scalar matrix such that (D - A(G)) becomes a diagonally dominant matrix.