• Journal of Korean Society for Quality Management
• /
• v.25 no.4
• /
• pp.106-114
• /
• 1997

The smoothing parameter or bandwidth is crucial to performance of the kernel based regression estimator. So the choice of a "optimal" smoothing parameter produce a single curve estimate. If a single estimate is replaced by a family of estimates, it become easy that we understand what varies with choice of the smoothing parameter. This paper suggests the threshold of the maximum bandwidth and the number of the family members in the regression context.n context.

• Communications for Statistical Applications and Methods
• /
• v.2 no.2
• /
• pp.347-349
• /
• 1995

In this paper a central limit theorem on $L^P(R)$ for $1{\leq}p<{\infty}$ is obtained with an example when ${X_n}$ is a sequence of independent, identically distributed random variables on $L^P(R)$.

• Journal of the Korean Statistical Society
• /
• v.24 no.2
• /
• pp.349-360
• /
• 1995

This paper deals with the robustness properties of the minimum disparity estimation in linear regression models. The estimators defined as statistical quantities whcih minimize the blended weight Hellinger distance between a weighted kernel density estimator of the residuals and a smoothed model density of the residuals. It is shown that if the weights of the density estimator are appropriately chosen, the estimates of the regression parameters are robust.

• Proceedings of the Korean Information Science Society Conference
• /
• 1998.10b
• /
• pp.188-190
• /
• 1998

데이터 베이스 관리 시스템에서는 질의 결과의 크기(selectivity)를 미리 예측하는 것이 필요하다. 질의 결과의 크기는 데이터의 분포 상태에 의해서 결정된다. 이러한 데이터의 분포 상태를 정확하게 예측하는 것이 매우 중요하다. 대부분의 데이터 베이스 관리 시스템에서는 이를 위하여 주기적으로 저장하고 있는 레코드에 대해서 히스토그램을 만들고 이용한다. 이 방법은 히스토그램의 저장공간이 적게 필요로 하고 선택도를 추정하는데 있어서 선택도 추정시 부가적인 계산이 필요하지 않은 장점이 있지만, 일정한 크기의 버켓내에서는 데이터들이 균일하게 분포한다는 가정을 함으로써 선택도 추정에 있어서 에러율이 높았다. 이에 본 논문에서는 커널 방법을 사용하여 버켓 내 데이터의 분포에 대하여 추정 함으로써 이를 해결하는 방법을 제시하였다.

• Journal of the Korean Data and Information Science Society
• /
• v.10 no.1
• /
• pp.111-118
• /
• 1999

We treat the problem of estimating reliability R(t) = P[Y(t) < X(t)] in the time dependent strength-stress model in which a unit of strength X(t) is subjected to environmental stress Y(t) at time t. In this paper two nonparametric approaches to estimate of R(t) are analyzed and compared with parametric method by simulation.

• Computers and Concrete
• /
• v.4 no.4
• /
• pp.299-316
• /
• 2007

The compressive strength of concrete is a commonly used criterion in producing concrete. However, the test on the compressive strength is complicated and time-consuming. More importantly, since the test is usually performed 28 days after the placement of the concrete at the construction site, it is too late to make improvements if unsatisfactory test results are incurred. Therefore, an accurate and practical strength estimation method that can be used before the placement of concrete is highly desirable. In this study, the estimation of the concrete strength is performed using support vector regression (SVR) based on the mix proportion data from two ready-mixed concrete companies. The estimation performance of the SVR is then compared with that of neural network (NN). The SVR method has been found to be very efficient in estimation accuracy as well as computation time, and very practical in terms of training rather than the explicit regression analyses and the NN techniques.

• Communications for Statistical Applications and Methods
• /
• v.16 no.3
• /
• pp.519-528
• /
• 2009

This paper deals with nonparametric estimation of discontinuous regression curve. Quite number of researches about this topic have been done. These researches are classified into two categories, the indirect approach and direct approach. The major goal of the indirect approach is to obtain good estimates of jump locations, whereas the major goal of the direct approach is to obtain overall good estimate of the regression curve. Thus it seems that two approaches are quite different in nature, so people say that the comparison of two approaches does not make much sense. Therefore, a thorough comparison of them is lacking. However, even though the main issue of the indirect approach is the estimation of jump locations, it is too obvious that we have an estimate of regression curve as the subsidiary result. The point is whether the subsidiary result of the indirect approach is as good as the main result of the direct approach. The performance of two approaches is compared through a simulation study and it turns out that the indirect approach is a very competitive tool for estimating discontinuous regression curve itself.

• Communications for Statistical Applications and Methods
• /
• v.27 no.5
• /
• pp.547-568
• /
• 2020

Modeling the statistical autocorrelations in spatial data is often achieved through the estimation of the variograms, where the selection of the appropriate valid variogram model, especially for small samples, is crucial for achieving precise spatial prediction results from kriging interpolations. To estimate such a variogram, we traditionally start by computing the empirical variogram (traditional Matheron or robust Cressie-Hawkins or kernel-based nonparametric approaches). In this article, we conduct numerical studies comparing the performance of these empirical variograms. In most situations, the nonparametric empirical variable nearest-neighbor (VNN) showed better performance than its competitors (Matheron, Cressie-Hawkins, and Nadaraya-Watson). The analysis of the spatial groundwater dataset used in this article suggests that the wave variogram model, with hole effect structure, fitted to the empirical VNN variogram is the most appropriate choice. This selected variogram is used with the ordinary kriging model to produce the predicted pollution map of the nitrate concentrations in groundwater dataset.

• 한국연소학회:학술대회논문집
• /
• 2013.06a
• /
• pp.19-20
• /
• 2013

The cofiring of renewable fuel in coal fired boilers is an attractive option to mitigate $CO_2$ emissions, since it is relatively low cost option for efficiently converting renewable fuel to electricity by adding biomass as partial substitute of coal. However, it would lead to reduce plant efficiency and flexibility in operation, and increase operation cost and capital cost associated with renewable fuels handling and firing equipment. The aim of this study is to investigate reduction of carbon dioxide at varying percentage of biomass in fuel blend to the boiler biomass, and estimate operation and capital cost. Wood pellet, PKS (palm kernel shell), EFB (empty fruit bunch) and sludge are considered as a renewable fuels for a cofiring with coal. Several approaches by the cofiring ratio are chosen from past plant demonstrations and commercial cofiring operation, and they are evaluated and discussed for CO2 reduction and cost estimation.

• Communications for Statistical Applications and Methods
• /
• v.29 no.1
• /
• pp.127-150
• /
• 2022

We discuss multiple change-point estimation as edge detection in piecewise smooth functions with finitely many jump discontinuities. In this paper we propose change-point estimators using concentration kernels with Fourier coefficients. The change-points can be located via the signal based on Fourier transformation system. This method yields location and amplitude of the change-points with refinement via concentration kernels. We prove that, in an appropriate asymptotic framework, this method provides consistent estimators of change-points with an almost optimal rate. In a simulation study the proposed change-point estimators are compared and discussed. Applications of the proposed methods are provided with Nile flow data and daily won-dollar exchange rate data.