• Proceedings of the Korean Statistical Society Conference
• /
• 2002.11a
• /
• pp.103-108
• /
• 2002

We consider an estimation of discontinuous variance function in nonparametric heteroscedastic random design regression model. We first propose estimators of a change point and jump size in variance function and then construct an estimator of entire variance function. We examine the rates of convergence of these estimators and give results on their asymptotics. Numerical work reveals that the effectiveness of change point analysis in variance function estimation is quite significant.

• Journal of the Korean Statistical Society
• /
• v.35 no.1
• /
• pp.1-23
• /
• 2006

In this paper we consider an estimation of the discontinuous variance function in nonparametric heteroscedastic random design regression model. We first propose estimators of the change point in the variance function and then construct an estimator of the entire variance function. We examine the rates of convergence of these estimators and give results for their asymptotics. Numerical work reveals that using the proposed change point analysis in the variance function estimation is quite effective.

• Journal of the Korean Statistical Society
• /
• v.30 no.2
• /
• pp.265-280
• /
• 2001

Estimation of the edge of a distribution has many important applications. It is related to classification, cluster analysis, neural network, and statistical image recovering. The problem also arises in measuring production efficiency in economic systems. Three most promising nonparametric estimators in the existing literature are introduced. Their statistical properties are provided, some of which are new. Themes of future study are also discussed.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.11 no.1
• /
• pp.73-84
• /
• 1986

A statistical method is proposed which estimates an edge that is one of the basic features in image understanding. The conventional edge detection techniques are performed well for a deterministic singnal, but are not satisfactory for a statistical signal. In this paper, we use the likelihood function which takes account of the statistical property of a signal, and derive the decision function from it. We propose the maximum likelihood edge detection technique which estimates an edge point which maximizes the decision function mentioned above. We apply this technique to statistecal signals which are generated by using the random number generator. Simnulations show that the statistical edge detection technique gives satisfactory results. This technique is extended to the two-dimensional image and edges are found with a good accuracy.

• Communications for Statistical Applications and Methods
• /
• v.1 no.1
• /
• pp.20-26
• /
• 1994

The hazard function is a non-negative function that measures the propensity of failure in the immediate furture, and is frequently used as a decision criterion, especially in replacement decisions. In this paper, we compute approximate confidence intervals for the lognormal hazard function under Type I censored data, and show how to choose the sample size needed to estimate a point on the hazard function with a specified degree of precision. Also we provide a table that can be used to compute the sample size.

• The Journal of Fisheries Business Administration
• /
• v.23 no.2
• /
• pp.91-99
• /
• 1992

It has been one of the critical issues that the researchers properly evaluate the fishing rights foregone by the coastal development activities like wetland reclamation. Particularly, estimating the productivity of concerned fishing rights is known to play a significant role in their monetary compensation. As a result, this paper attempts to develop a statistical model characterized by Cobb-Douglas production function in conjunction with the fisheries' productivity estimation. The primary hypotheses involving their statistical production function are as below : 1. The quantity of fisheries production is hypothesized to be expressed as a function of capital (K) and labor(L) put into fishing activities. 2. The estimated parameters of K and L are hypothesized to satisfy the conventional condition of production function as a form of Cobb-Douglas. These statistical tests reveal that the shellfish farming productivity heavily depends on the acre of mariculture while the input of labor force also considerably affects its productivity. In case of the fixed net fishing productivity, both the factors of capital and labor similarly affect the marginal change in its productivity. En addition, the productivity of shellfish (arming turns out to follow the increasing returns to scale, whereas that of fixed net fishing comes up with the decreasing returns to scale.

• Transactions of the Korean Society of Mechanical Engineers
• /
• v.14 no.5
• /
• pp.1290-1299
• /
• 1990

Characteristics of flow patterns in horizontal gas-liquid two-phase flow for two different sizes of pipe were investigated based upon a statistical analysis of differential pressure fluctuations at an orifice. The probability density function and the power spectral density function of the traces indicate peculiar shapes depending upon the two-phase flow regime. Mixed and separated flows also could be identified by the autocorrelation function. The transition region from separated flow to mixed flow also could be identified by these statistical properties. The experimental data determined by this method were compared with the flow pattern maps suggested by other investigators. The result indicates that the statistical characteristics of differential pressure fluctuations at orifices may be a useful tool for identifying flow patterns of horizontal gas-liquid two-phase flow.

• Communications for Statistical Applications and Methods
• /
• v.15 no.5
• /
• pp.783-791
• /
• 2008

Estimation procedure of the finite population proportion and distribution function is considered. Based on a logistic regression model, an approximately model- optimal estimator is defined and conditions for the estimator to be design-consistent are given. Simulation study shows that the model-optimal design-consistent estimator defined under a logistic regression model performs well in estimating the finite population distribution function.

• Communications for Statistical Applications and Methods
• /
• v.10 no.2
• /
• pp.497-505
• /
• 2003

Under normality assumption, the tolerance interval for a future observation is sometimes of great interest in statistics. In this paper, we state the influence function on the standard deviation $\sigma$, and use it to derive the influence function on tolerance limits. Simulation study shows that the two influence functions perform very well.

• Communications for Statistical Applications and Methods
• /
• v.10 no.3
• /
• pp.1025-1036
• /
• 2003

Variable selection algorithm for principal component analysis using penalty function is proposed. We use the fact that usual principal component problem can be expressed as a maximization problem with appropriate constraints and we will add penalty function to this maximization problem. Simulated annealing algorithm is used in searching for optimal solutions with penalty functions. Comparisons between several well-known penalty functions through simulation reveals that the HARD penalty function should be suggested as the best one in several aspects. Illustrations with real and simulated examples are provided.