• Journal of the Korean Data and Information Science Society
• /
• 제22권6호
• /
• pp.1223-1232
• /
• 2011

Tong and Wang's estimator (2005) is a new approach to estimate the error variance using least squares method such that a simple linear regression is asymptotically derived from Rice's lag- estimator (1984). Their estimator highly depends on the setting of a regressor and weights in small sample sizes. In this article, we propose a new approach via a local quadratic approximation to set regressors in a small sample case. We estimate the error variance as the intercept using a ridge regression because the regressors have the problem of multicollinearity. From the small simulation study, the performance of our approach with some existing methods is better in small sample cases and comparable in large cases. More research is required on unequally spaced points.

• Journal of the Korean Statistical Society
• /
• 제18권1호
• /
• pp.46-61
• /
• 1989

This paper is concerned with the development of a nonparametric procedure for the statistical inference about the nonlinear regression parameters. A confidence region and a hypothesis testing procedure based on a class of signed linear rank statistics are proposed and the asymptotic distributions of the test statistic both under the null hypothesis and under a sequence of local alternatives are investigated. Some desirable asymptotic properties including the asymptotic relative efficiency are discussed for various score functions.

• Communications for Statistical Applications and Methods
• /
• 제4권1호
• /
• pp.177-183
• /
• 1997

A data-adaptive order selection procedure is proposed for local polynomial nonparametric regression. For each given polynomial order, bias and variance are estimated and the adaptive polynomial order that has the smallest estimated mean squared error is selected locally at each location point. To estimate mean squared error, empirical bias estimate of Ruppert (1995) and local polynomial variance estimate of Ruppert, Wand, Wand, Holst and Hossjer (1995) are used. Since the proposed method does not require fitting polynomial model of order higher than the model order, it is simpler than the order selection method proposed by Fan and Gijbels (1995b).

• Communications for Statistical Applications and Methods
• /
• 제8권1호
• /
• pp.1-14
• /
• 2001

This paper discusses choosing the weight function of the Hardle and Mammen statistic in nonparametric goodness-of-fit test for regression curve. For this purpose, we modify the Hardle and Mammen statistic and derive its asymptotic distribution. Some results on the test statistic from the wild bootstrapped sample are also obtained. Through Monte Carlo experiment, we check the validity of these results. Finally, we study the powers of the test and compare with those of the Hardle and Mammen test through the simulation.

• Journal of the Korean Statistical Society
• /
• 제26권1호
• /
• pp.89-100
• /
• 1997

In this paper we are concerned with comparing ordered treatment effects with a control in a randomized block design with multiple observations per cell. Two nonparametric procedures for detecting which treatment are better than the control are proposed and compared. An example is given and the results of a Monte Carlo power study are discussed.

• Communications for Statistical Applications and Methods
• /
• 제15권5호
• /
• pp.709-717
• /
• 2008

The difference of two one-sided kernel estimators is usually used to detect the location of the discontinuity points of regression function. The large absolute value of the statistic imply discontinuity of regression function, so we may use the difference of two one-sided kernel estimators as the test statistic for testing null hypothesis of a smooth regression function. The problem is, however, we only know the asymptotic distribution of the test statistic under $H_0$ and we hardly expect the good performance of test if we rely solely on the asymptotic distribution for determining the critical points. In this paper, we show that if we adjust the bias of test statistic properly, the asymptotic rules hold for even small sample size situation.

• Journal of the Korean Statistical Society
• /
• 제35권1호
• /
• pp.105-114
• /
• 2006

In this paper we propose a simple and computationally attractive difference-based variance estimator in nonparametric regression models with multivariate predictors. We show that the estimator achieves $n^{-1/2}$ rate of convergence for regression functions with only a first derivative when d, the dimension of the predictor, is less than or equal to 4. When d > 4, the rate turns out to be $n^{-4/(d+4)}$ under the first derivative condition for the regression functions. A numerical study suggests that the proposed estimator has a good finite sample performance.

• Journal of the Korean Statistical Society
• /
• 제25권4호
• /
• pp.499-514
• /
• 1996

Consider an additive regression model of Y on X = (X$_1$,X$_2$,. . .,$X_p$), Y = $sum_{j=1}^pf_j(X_j) + $\varepsilon$$, where $f_j$s are smooth functions to be estimated and $\varepsilon$ is a random error. If $X_j$s are fixed design points, we call it the fixed design additive model. Since the response variable Y is observed at fixed p-dimensional design points, the behavior of the nonparametric regression estimator depends on the design. We propose a fixed design called permutation fixed design, and fit the regression function by the kernel method. The estimator in the permutation fixed design achieves the univariate optimal rate of convergence in mean squared error for any p $\geq$ 2.

• Communications for Statistical Applications and Methods
• /
• 제19권6호
• /
• pp.885-898
• /
• 2012

In this paper, we consider Bayesian approaches to partially linear models, in which a regression function is represented by a semiparametric additive form of a parametric linear regression function and a nonparametric regression function. We make a comparative study on the performance of widely used Bayesian partially linear models in terms of empirical analysis. Specifically, we deal with three Bayesian methods to estimate the nonparametric regression function, one method using Fourier series representation, the other method based on Gaussian process regression approach, and the third method based on the smoothness of the function and differencing. We compare the numerical performance of three methods by the root mean squared error(RMSE). For empirical analysis, we consider synthetic data with simulation studies and real data application by fitting each of them with three Bayesian methods and comparing the RMSEs.

• 환경영향평가
• /
• 제4권2호
• /
• pp.93-103
• /
• 1995

This study was carried out to suggest the best statistical test to analyze the trend in monthly water quality data. Traditional parametric tests such as t-test and regression analysis are based on the assumption that the underlying population has a normal distribution and regression analysis additionally assumes that residual errors are independent. Analyzing 9-years monthly COD data collected at Paldang in Han River, the underlying population was found to be neither normal nor independent. Therefore parametric tests are invalid for trend detection. Four Kinds of nonparametric statistical tests, such as Run Test, Daniel test, Mann-Kendall test, and Time Series Residual Analysis were applied to analyze the trend in the COD data, Daniel test and Mann-Kendall test indicated upward trend in COD data. The best nonparametric test was suggested to be Daniel test, which is simple in computation and easy to understand the intuitive meaning.