• Communications for Statistical Applications and Methods
• /
• v.29 no.1
• /
• pp.65-83
• /
• 2022

Although various statistical methods have been developed to map time-dependent genetic factors, most identified genetic variants can explain only a small portion of the estimated genetic variation in longitudinal traits. Gene-gene and gene-time/environment interactions are known to be important putative sources of the missing heritability. However, mapping epistatic gene-gene interactions is extremely difficult due to the very large parameter spaces for models containing such interactions. In this paper, we develop a Gaussian process (GP) based nonparametric Bayesian variable selection method for longitudinal data. It maps multiple genetic markers without restricting to pairwise interactions. Rather than modeling each main and interaction term explicitly, the GP model measures the importance of each marker, regardless of whether it is mostly due to a main effect or some interaction effect(s), via an unspecified function. To improve the flexibility of the GP model, we propose a novel grid-based method for the within-subject dependence structure. The proposed method can accurately approximate complex covariance structures. The dimension of the covariance matrix depends only on the number of fixed grid points although each subject may have different numbers of measurements at different time points. The deviance information criterion (DIC) and the Bayesian predictive information criterion (BPIC) are proposed for selecting an optimal number of grid points. To efficiently draw posterior samples, we combine a hybrid Monte Carlo method with a partially collapsed Gibbs (PCG) sampler. We apply the proposed GP model to a mouse dataset on age-related body weight.

• 한국전산유체공학회:학술대회논문집
• /
• 2011.05a
• /
• pp.194-198
• /
• 2011

In this study, an efficient algorithm for Delaunay triangulation of a number of points which can be used on a GPU-based parallel computation is studied The developed algorithm is programmed using CUDA library. and the program takes full advantage of parallel computation which are concurrently performed on each of the threads on GPU. The results of partitioned triangulation collected from the GPU computation requires proper stitching between neighboring partitions and calculation of connectivities among triangular cells on CPU In this study, the effect of number of threads on the efficiency and total duration for Delaunay grid generation is studied. And it is also shown that GPU computing using CUDA for Delaunay grid generation is feasible and it saves total time required for the triangulation of the large number points compared to the sequential CPU-based triangulation programs.

• 한국전산유체공학회:학술대회논문집
• /
• 1999.11a
• /
• pp.108-114
• /
• 1999

In this paper. a new automatic multi=block grid generation technique for general 2D regions is introduced. According to this simple and robust method, the domain of interest is first triangulated by using Delaunay triangulation of boundary points, and then geometric information of those triangles is used to obtain block topology. Once block boundaries are obtained. structured grid for each block is generated such that grid lines have $C^0-continuity$ across inter-block boundaries. In the final step of the present method, an elliptic grid generation method is applied to smoothen grid distribution for each block and also to re-locale the inter-block boundaries, and eventually to achieve a globally smooth multi-block structured grid system with $C^1-continuity$.

• Journal of the Korean Society of Groundwater Environment
• /
• v.6 no.1
• /
• pp.23-32
• /
• 1999

A new numerical method using solution-adaptive grids (SAG) is developed to solve the Richards' equation (RE) for unsaturated flow in porous media. Using a grid generation technique, the SAG method automatically redistributes a fixed number of grid points during the flow process, so that more grid points are clustered in regions of large solution gradients. The method uses the coordinate transformation technique to employ a new transformed RE, which is solved with the standard finite difference method. The movement of grid points is incorporated into the transformed RE, and therefore all computation is performed on fixed grid points of the transformed domain without using any interpolation techniques. Thus, numerical difficulties arising from the movement of the wetting front during the infiltration process have been substantially overcome by the new method. Numerical experiments for an one-dimensional infiltration problem are presented to compare the SAG method to the modified Picard method using a fixed grid. Results show that accuracy of a SAG solution using 41 nodes is comparable with the solution of the fixed grid method using 201 nodes, while it requires only 50% of the CPU time. The global mass balance and the convergence of SAG solutions are strongly affected by the time step size (Δt) and the weighting parameter (${\gamma}$) used for generating solution-adaptive grids. Thus, the method requires automated readjustment of Δt and ${\gamma}$ to yield mass-conservative and convergent solutions, although it may increase computational costs. The method can be effective especially for simulating unsaturated flow and other transport problems involving the propagation of a sharp-front.

• Proceedings of the Korean Society of Surveying, Geodesy, Photogrammetry, and Cartography Conference
• /
• 2007.04a
• /
• pp.179-182
• /
• 2007

In this study, we proposed a way to detect trees using virtual grid and to discriminate vertical wall points from building tops based on effective segmentation of LIDAR point cloud utilizing scan line characteristics. Trees were detected by their surface roughness value calculated based on virtual grid and vertical building wall points were discriminated from building tops with one dimensional filtering of scan line during segmenting point cloud. In results, we could distinguish trees from buildings and bind virtical wall points to prevent them from faltly acting on point segmentation process.

• Journal of the Society of Naval Architects of Korea
• /
• v.39 no.3
• /
• pp.18-27
• /
• 2002

In order to improve the accuracy of the solution near the boundary in an analysis of viscous flow around an arbitrary boundary which move and be deformed using an Eulerian concept, a level-set based grid deformation method is introduced to concentrate grid points near the boundary. This paper presents a new monitor function which can easily control the level of the concentration of grid points along the boundary. Computations for steady flow around a semi-circular cylinder mounted on the bottom of the flow domain were carried out to check the improvement of the solution using the adaptive grid system with an immersed boundary method. The present numerical results show a good agreement with the solutions obtained by a body fitted grid system and more accurate solutions than those computed with non-adaptive grid system. For the validation of mechanical usefulness of the present method, an expanded analysis of flow around multi-body fixed in the flow domain was carried out. Finally, the present moving adaptive grid method was applied to a two-dimensional bubble rise problem. The computed results show well adapted grid points around the boundary of the bubble at every time and a good agreement with the result calculated by fixed grid system.

• Journal of computational fluids engineering
• /
• v.6 no.4
• /
• pp.15-25
• /
• 2001

Numerical prediction of the diffusion controlled transition in a turbine gas pass is important because it can change the local heat transfer rate over a turbine blade as much as three times. In this study, the gas flow over turbine blade is simplified to the flat plate boundary layer, and an adaptive grid scheme redistributing grid points within the computation domain is proposed with a great emphasis on the construction of the grid control function. The function is sensitized to the second invariant of the mean strain tensor, its spatial gradient, and the interaction of pressure gradient and flow deformation. The transition process is assumed to be described with a κ-ε turbulence model. An elliptic solver is employed to integrate governing equations. Numerical results show that the proposed adaptive grid scheme is very effective in obtaining grid independent numerical solution with a very low grid number. It is expected that present scheme is helpful in predicting actual flow within a turbine to improve computation efficiency.

• 한국전산유체공학회:학술대회논문집
• /
• 1999.11a
• /
• pp.145-151
• /
• 1999

An Interactive grid generation program(KGRID) with graphical user interface(GUI) has been improved. KGRID works on the UNLX environment and GUI has been implemented with OSF/Motif and X Toolkit and the graphics language is Open GL for visualization of the 3D objects. It supports more convenient user environment to generate 2D and 3D multi-block structured grid systems. It provides various useful field grid generation methods, which are the algebraic methods, the elliptic partial differential equations method and the predictor-corrector method. It also supports 3D surface grid generation with NURBS(Non-Uniform Rational B-Spline) and various stretching functions to control grid points distribution on curves and surfaces. And some menus are added to perform flexible management, for the objects. We generated surface and field grid system about full aircraft configuration using KGRID. The performance and stability of the KGRID is verified through the generation of the grid system about a complex shape.

• Nuclear Engineering and Technology
• /
• v.39 no.6
• /
• pp.737-746
• /
• 2007

A spacer grid, which is one of the most important structural components in a PWR fuel assembly, supports its fuel rods laterally and vertically. Based on in-house design experience, scrutiny of the design features of advanced nuclear fuels and the patents of other spacer grids, KAERI has devised its own spacer grid shapes and acquired patents. In this study, a performance evaluation of KAERI's spacer grid shapes was carried out from mechanical/structural and thermohydraulic view points. A comparative performance evaluation of commercial spacer grid shapes was also carried out. The comparisons addressed the spring characteristics, fuel rod vibration characteristics, fretting wear resistance, impact strength characteristics, CHF enhancement, and the pressure drop level of the spacer grid shapes. The results show that the performances of KAERI's spacer grid shapes are as good as or better than those of the commercial spacer grid shapes.

• Journal of applied mathematics & informatics
• /
• v.30 no.1_2
• /
• pp.235-244
• /
• 2012

In this paper, we solve the three-dimensional biharmonic equation with Dirichlet boundary conditions of second kind using the full multigrid (FMG) algorithm. We derive a finite difference approximations for the biharmonic equation on a 18 point compact stencil. The unknown solution and its second derivatives are carried as unknowns at grid points. In the multigrid methods, we use a fourth order interpolation to producing a new intermediate unknown functions values on a finer grid, and the full weighting restriction operators to calculating the residuals at coarse grid points. A set of test problems gives excellent results.