• Structural Engineering and Mechanics
• /
• v.92 no.1
• /
• pp.111-120
• /
• 2024

Sampling methods are powerful approaches to solving the problems of structural reliability analysis and estimating the failure probability of structures. In this paper, a new sampling method is proposed offering lower variance and lower computational cost for complex and high-dimensional problems. The method is called Optimal Latinized partially stratified sampling (OLPSS) as it is based upon the Latinized Partially Stratified Sampling (LPSS) which itself is based on merging Stratified Sampling (SS) and Latin Hypercube Sampling (LHS) algorithms. While LPSS has a low variance, it may suffer from a lack of good space-filling of its generated samples in some cases. In the OLPSS, this issue has been resolved by employing a new columnwise-pairwise exchange optimization procedure for sample generation. The efficiency of the OLPSS has been tested and reported under several benchmark mathematical functions and structural examples including structures with a large number of variables (e.g., a structure with 67 variables). The proposed method provides highly accurate estimates of the failure probability of structures with a significantly lower variance relative to the Monte Carlo simulations, Latin Hypercube, and standard LPSS.

• Bulletin of the Korean Mathematical Society
• /
• v.55 no.4
• /
• pp.1303-1315
• /
• 2018

For principal component analysis (PCA) to efficiently analyze large scale matrices, it is crucial to find a few singular vectors in cheaper computational cost and under lower memory requirement. To compute those in a fast and robust way, we propose a new stochastic method. Especially, we adopt the stochastic variance reduced gradient (SVRG) method [11] to avoid asymptotically slow convergence in stochastic gradient descent methods. For that purpose, we reformulate the PCA problem as a unconstrained optimization problem using a quadratic penalty. In general, increasing the penalty parameter to infinity is needed for the equivalence of the two problems. However, in this case, exact penalization is guaranteed by applying the analysis in [24]. We establish the convergence rate of the proposed method to a stationary point and numerical experiments illustrate the validity and efficiency of the proposed method.

• Communications for Statistical Applications and Methods
• /
• v.6 no.2
• /
• pp.487-500
• /
• 1999

We propose a class of hierarchical Bayes estimators of the error variance under the relative squared error loss in balanced fixed-effects two-way analysis of variance models. Also we provide analytic expressions for the risk improvement of the hierarchical Bayes estimators over multiples of the error sum of squares. Using these expressions we identify a subclass of the hierarchical Bayes estimators each member of which dominates the best multiple of the error sum of squares which is known to be minimax. Numerical values of the percentage risk improvement are given in some special cases.

• Journal of the Korean Data and Information Science Society
• /
• v.17 no.2
• /
• pp.559-568
• /
• 2006

In this paper, we develop the noninformative priors for the linear mixed models when the parameter of interest is the ratio of variance components. We developed the first and second order matching priors. We reveal that the one-at-a-time reference prior satisfies the second order matching criterion. It turns out that the two group reference prior satisfies a first order matching criterion, but Jeffreys' prior is not first order matching prior. Some simulation study is performed.

• East Asian mathematical journal
• /
• v.34 no.3
• /
• pp.239-248
• /
• 2018

In this paper, we derive a closed-form formula of a second-order approximation for a European corrected option price under stochastic elasticity of variance model mentioned in Kim et al. (2014) [1] [J.-H. Kim, J Lee, S.-P. Zhu, S.-H. Yu, A multiscale correction to the Black-Scholes formula, Appl. Stoch. Model. Bus. 30 (2014)]. To find the explicit-form correction to the option price, we use Mellin transform approaches.

• Communications for Statistical Applications and Methods
• /
• v.12 no.1
• /
• pp.47-59
• /
• 2005

Expected mean square(EMS) is an important part of conducting the analysis of variance in the experimental design problem, especially in mixed or random models. We present a set of EMS rules for balanced factorial designs under no restriction on interaction. Also we point out how to use the variance component of SPSS or SAS procedure to obtain EMS.

• Journal of the Korean Statistical Society
• /
• v.9 no.1
• /
• pp.97-108
• /
• 1980

Variance components model appears often in designing experiments including time series data analysis. This paper is investigating the properties of the various procedures in estimating variance components for the two-way random model without interaction under normality. In this age of computer-oriented computations, MINQUE is found to be quite practicla because of the robustness with respect to the design configurations and parameters. Also adjusted AOV type estimation procedure is found to yield superior results over the unadjusted one.

• Journal of the Korea Computer Industry Society
• /
• v.9 no.5
• /
• pp.207-216
• /
• 2008

Statistical analysis menus of MS Excel software, with powerful spreadsheet functions has not been modified since Excel 2000 Edition and its utilization is very low. To improve utilization of Excel menu for statistical analysis, this research compared outputs of Excel statistical menus and computed test statistics, and developed high-level macros. Outputs of Excel menus, both oneway layout and twoway layout, on real data are exactly same as the computed test statistics, and therefore, Excel menus for statistical analysis are reliable. Macros to provide results for Analysis of Variance with a block and multiple comparison of means are developed using Excel functions.

• Journal of Information Processing Systems
• /
• v.16 no.4
• /
• pp.832-844
• /
• 2020

The degree of dispersion of a random variable can be described by the variance, which reflects the distance of the random variable from its mean. However, the time complexity of the traditional variance calculation algorithm is O(n), which results from full calculation of all samples. When the number of samples increases or on the occasion of high speed signal processing, algorithms with O(n) time complexity will cost huge amount of time and that may results in performance degradation of the whole system. A novel multi-step recursive algorithm for variance calculation of the time-varying data series with O(1) time complexity (constant time) is proposed in this paper. Numerical simulation and experiments of the algorithm is presented and the results demonstrate that the proposed multi-step recursive algorithm can effectively decrease computing time and hence significantly improve the variance calculation efficiency for time-varying data, which demonstrates the potential value for time-consumption data analysis or high speed signal processing.

• The Korean Journal of Applied Statistics
• /
• v.25 no.1
• /
• pp.55-66
• /
• 2012

Option pricing models using L$\acute{e}$evy processes are suggested as an alternative to the Black-Scholes model since empirical studies showed that the Black-Sholes model could not reflect the movement of underlying assets. In this paper, we investigate whether the Variance Gamma model can reflect the movement of underlying assets in the Korean stock market better than the Black-Scholes model. For this purpose, we estimate parameters and perform likelihood ratio tests using KOSPI 200 data based on the density for the log return and the option pricing formula proposed in Madan et al. (1998). We also calculate some statistics to compare the models and examine if the volatility smile is corrected through regression analysis. The results show that the option price estimated under the Variance Gamma process is closer to the market price than the Black-Scholes price; however, the Variance Gamma model still cannot solve the volatility smile phenomenon.