• Journal of the Korean Data and Information Science Society
• /
• v.24 no.3
• /
• pp.533-541
• /
• 2013

In this study, a simple principal component analysis using Lasso is proposed. This method consists of two steps. The first step is to compute principal components by the principal component analysis. The second step is to regress each principal component on the original data matrix by Lasso regression method. Each of new principal components is computed as the linear combination of original data matrix using the scaled estimated Lasso regression coefficient as the coefficients of the combination. This method leads to easily interpretable principal components with more 0 coefficients by the properties of Lasso regression models. This is because the estimator of the regression of each principal component on the original data matrix is the corresponding eigenvector. This method is applied to real and simulated data sets with the help of an R package for Lasso regression and its usefulness is demonstrated.

• Structural Engineering and Mechanics
• /
• v.30 no.5
• /
• pp.619-645
• /
• 2008

This study presents the efficiency of simulating structural systems using a method that combines a simplified component model (SCM) and rigorous component model (RCM). To achieve a realistic simulation of structural systems, a numerical model must be adequately capturing the detailed behaviors of real systems at various scales. However, capturing all details represented within an entire structural system by very fine meshes is practically impossible due to technological limitations on computational engineering. Therefore, this research develops an approach to simulate large-scale structural systems that combines a simplified global model with multiple detailed component models adjusted to various scales. Each correlated multi-scale simulation model is linked to others using a multi-level hierarchical modeling simulation method. Simulations are performed using nonlinear finite element analysis. The proposed method is applied in an analysis of a simple reinforced concrete structure and the Reuipu Elementary School (an existing structure), with analysis results then compared to actual onsite observations. The proposed method obtained results very close to onsite observations, indicating the efficiency of the proposed model in simulating structural system behavior.

• Proceedings of the Korean Statistical Society Conference
• /
• 2003.05a
• /
• pp.151-156
• /
• 2003

Variable selection algorithm for principal component analysis using penalized likelihood method is proposed. We will adopt a probabilistic principal component idea to utilize likelihood function for the problem and use HARD penalty function to force coefficients of any irrelevant variables for each component to zero. Consistency and sparsity of coefficient estimates will be provided with results of small simulated and illustrative real examples.

• Journal of Korea Game Society
• /
• v.1 no.1
• /
• pp.73-79
• /
• 2001

The objective of this paper is to systematically establish the development of game software through the fusion of game implementation and component methodology. The development of game software has inconsistently accomplished without the consistent frame for the analysis and design of game development in the domestic environments. Therefore in case of the development of game software based on game component, although a lot of resources were required in the initial the accumulation of experience and the technology of reuse lead to the efficiency of the maintenance ultimately.

• Proceedings of the Earthquake Engineering Society of Korea Conference
• /
• 1997.10a
• /
• pp.151-158
• /
• 1997

The main concern of numerical dynamic analysis of large structures is to find an acceptable solution with fewer mode shapes and less computational efforts. component mode method utilizes substructure technique to reduce the degrss of freedom but have a disadvantage to not consider the dynamic characteristics of loads. Ritz Vector method consider the load characteristics but requires many integrations and errors are accumulated. In this study, to prove the effectiveness of component mode method, Lanczos algorithm are introduced. To prove the effectiveness of this method, example structures areanalyzed and the results are compared with SAP90.

• Computational Structural Engineering
• /
• v.9 no.2
• /
• pp.115-120
• /
• 1996

The main concern of numerical dynamic analysis of large structures is to find an acceptable solution with fewer mode shapes and less computational efforts. Component mode method utilizes substructure technique to reduce the degree of freedom but have a disadvantage to not consider the dynamic characteristics of loads. Ritz Vector method consider the load characteristics but requires many integrations and errors are accumulated. In this study, to improve the effectiveness of component mode method, Lanczos algorithm is introduced. To prove the effectiveness of this method, example structure are analyzed and the results are compared with SAP90.

• Proceedings of the KSME Conference
• /
• 2007.05a
• /
• pp.921-926
• /
• 2007

A dynamic system under random disturbance is considered in the study. In order to control the system efficiently, proper reduction of system dimension is indispensible in design stage. The reduction method using component cost analysis in conjunction with stochastic analysis is proposed for the control of a system. System response is obtained in terms of dynamic moment equation via Fokker-Plank-Kolmogorov(F-P-K) equation. The dynamic moment response of the system under random disturbance are reduced by using of deterministic version of component cost analysis. The reduced system via proposed "stochastic component cost analysis" is successfully implemented for dynamic response and shows remarkable control performance effectively utilizing "stochastic controller" in physical time domain.

• The Journal of the Korea Contents Association
• /
• v.6 no.1
• /
• pp.54-63
• /
• 2006

In this paper we will propose a domain analysis methodology that uses an extended workflow mechanism based on dynamic modeling to solve problems of a traditional domain analysis on legacy systems. This methodology is called WODA(Workflow Oriented Domain Analysis). Following procedures on WODA, we can identify common/uncommon component, and also extract the cluster of components. It will be effectively reusable on developing new systems with these components. With our proposed component testing metrics, we can determine highly reusable component/scenario on identifying possible scenarios of the particular system. We can also recognize most critical/most frequent reusable components and prioritize possible component scenarios of the system. This paper contains one application of UPS that illustrates our autonomous modeling tool, WODA.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.29 no.1C
• /
• pp.65-71
• /
• 2004

We improved the MLLR speaker adaptation algorithm with reduction of the order of HMM parameters using PCA(Principle Component Analysis) or ICA(Independent Component Analysis). To find a smaller set of variables with less redundancy, we adapt PCA(principal component analysis) and ICA(independent component analysis) that would give as good a representation as possible, minimize the correlations between data elements, and remove the axis with less covariance or higher-order statistical independencies. Ordinary MLLR algorithm needs more than 30 seconds adaptation data to represent higher word recognition rate of SD(Speaker Dependent) models than of SI(Speaker Independent) models, whereas proposed algorithm needs just more than 10 seconds adaptation data. 10 components for ICA and PCA represent similar performance with 36 components for ordinary MLLR framework. So, compared with ordinary MLLR algorithm, the amount of total computation requested in speaker adaptation is reduced by about 1/167 in proposed MLLR algorithm.

• The Korean Journal of Applied Statistics
• /
• v.30 no.1
• /
• pp.135-145
• /
• 2017

Principal component analysis (PCA) describes the variation of multivariate data in terms of a set of uncorrelated variables. Since each principal component is a linear combination of all variables and the loadings are typically non-zero, it is difficult to interpret the derived principal components. Sparse principal component analysis (SPCA) is a specialized technique using the elastic net penalty function to produce sparse loadings in principal component analysis. When data are structured by groups of variables, it is desirable to select variables in a grouped manner. In this paper, we propose a new PCA method to improve variable selection performance when variables are grouped, which not only selects important groups but also removes unimportant variables within identified groups. To incorporate group information into model fitting, we consider a hierarchical lasso penalty instead of the elastic net penalty in SPCA. Real data analyses demonstrate the performance and usefulness of the proposed method.