• Journal of the Korean Statistical Society
• /
• 제24권2호
• /
• pp.337-347
• /
• 1995

In the problem of some sequential estimation, the stopping times may be written in the form $N(c) = inf{n \geq n_0; n \geq c^2 S^2_n/\delta^2 (\bar{X}_n)}$ where ${s^2_n}$ and ${\bar{X}_n}$ are the sequences of sample variance and sample mean of the independently and identically distributed (i.i.d.) random variables with distribution $F_{\theta}(x), \theta \in \Theta$, respectively, and $\delta$ is either constant or any given positive real valued function. We obtain some asymptotic normality and asymptotic expectation of the N(c) in various limiting situations. Specially, uniform asymptotic normality and uniform asymptotic expectation of the N(c) are given.

• Communications for Statistical Applications and Methods
• /
• 제20권1호
• /
• pp.63-75
• /
• 2013

In this paper, we derive maximum likelihood estimators (MLEs) and approximate maximum likelihood estimators (AMLEs) of unknown parameters in a generalized half logistic distribution under Type-II hybrid censoring. We also obtain approximate confidence intervals using asymptotic variance and covariance matrices based on the MLEs and the AMLEs. As an illustration, we examine the validity of the proposed estimation using real data. Finally, we compare the proposed estimators in the sense of the mean squared error (MSE), bias, and length of the approximate confidence interval through a Monte Carlo simulation for various censoring schemes.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 한국방송공학회 2012년도 하계학술대회
• /
• pp.262-265
• /
• 2012

In this paper we propose a new deinterlacing method with block classification and motion vector smoothing. The proposed method classifies a block, then depending on the region it belongs to, the motion estimation or line averaging is applied. To classify a block its variance is calculated. Then, for those blocks that belong to simple non-texture region the line averaging is done while motion estimation is applied to complex region. The motion vector field is smoothed using median filter what yields more accurate interpolation. In the experiments for the subjective evaluation, the proposed method bas shown satisfying results in terms of computation time consumption and peak signal-to-noise ratio. Due to the simplicity of the algorithm computation time was drastically decreased.

• Journal of the Korea Society for Simulation
• /
• 제9권4호
• /
• pp.67-75
• /
• 2000

Often system analysts are interested in the estimation of percentile for system performance. For instance, in PERT network system, the percentile that the project. Typically the control variate method is used to reduce the variability of mean response using the correlation between the response and the control variates with a little additional cost during the course of simulation. In the same spirit, we apply this method to estimate the percentile of project completion time in PERT system, and evaluate the efficiency of the controlled estimator for its percentile.1 Simulation results indicate that the controlled estimators are more effective in reducing the variances of estimators than the simple estimators, however those tend to a little underestimate the percentiles for some critical values. We need more simulation experiments to examine such a kind of bias problem. We expect this research presents a step forward in the area of variance reduction techniques of stochastic simulation.

• Journal of the Korean Data and Information Science Society
• /
• 제15권2호
• /
• pp.477-483
• /
• 2004

One of the major paradigms for supervised learning in neural network community is back-propagation learning. The standard implementations of back-propagation learning are optimal under the assumptions of identical and independent Gaussian noise. In this paper, for regression function estimation, we introduce $\varepsilon-insensitive$ back-propagation learning algorithm, which corresponds to minimizing the least absolute error. We compare this algorithm with support vector machine(SVM), which is another $\varepsilon-insensitive$ supervised learning algorithm and has been very successful in pattern recognition and function estimation problems. For comparison, we consider a more realistic model would allow the noise variance itself to depend on the input variables.

• Korean Journal of Child Studies
• /
• 제27권3호
• /
• pp.169-187
• /
• 2006

A hierarchical linear model(HLM) provides advantages over existing traditional statistical methods (e.g., ordinary least squares regression, repeated measures analysis of variance, etc.) for analyzing multilevel/longitudinal data or diary methods. HLM can gauge a more precise estimation of lower-level effects within higher-level units, as well as describe each individual's growth trajectory across time with improved estimation. This article 1) provides scholars who study children and families with an overview of HLM (i.e., statistical assumptions, advantages/disadvantages, etc.), 2) provides an empirical study to illustrate the application of HLM, and 3) discusses the application of HLM to the study of children and families. In addition, this article provided useful information on available articles and websites to enhance the reader's understanding of HLM.

• Journal of Korea Society of Industrial Information Systems
• /
• 제8권4호
• /
• pp.94-100
• /
• 2003

In this paper, we study the estimation of reliability function for a parallel system with k dependent exponential components. We assume that the failure of one of the k components changes the life distribution of the remaining components. Also, we compare with Cramer-Rna lower bound for variances of the minimum variance unbiased estimator, and the mean square errors of the maximum likelihood estimator of reliability system with the through the Monte Carlo simulation.

• The Transactions of the Korean Institute of Electrical Engineers A
• /
• 제48권5호
• /
• pp.560-570
• /
• 1999

This paper formulates the suboptimal robust Kalman filtering problem into two coupled Linear Matrix Inequality (LMI) problems by applying Lyapunov theory to the augmented system which is composed of the state equation in the uncertain linear system and the estimation error dynamics. This formulations not only provide the sufficient conditions for the existence of the desired filter, but also construct the suboptimal robust Kalman filter. The proposed filter can guarantee the optimized upper bound of the estimation error variance for uncertain systems with parametric uncertainties in both the state and measurement matrices. In addition, this paper shows how the problem of finding the minimizing solution subject to Quadratic Matrix Inequality (QMI), which cannot be easily transformed into LMI using the usual Schur complement formula, can be successfully modified into a generic LMI problem.

• Earthquakes and Structures
• /
• 제5권3호
• /
• pp.359-378
• /
• 2013

The problem of identification of multi-component and (or) spatially varying earthquake support motions based on measured responses in instrumented structures is considered. The governing equations of motion are cast in the state space form and a time domain solution to the input identification problem is developed based on the Kalman and particle filtering methods. The method allows for noise in measured responses, imperfections in mathematical model for the structure, and possible nonlinear behavior of the structure. The unknown support motions are treated as hypothetical additional system states and a prior model for these motions are taken to be given in terms of white noise processes. For linear systems, the solution is developed within the Kalman filtering framework while, for nonlinear systems, the Monte Carlo simulation based particle filtering tools are employed. In the latter case, the question of controlling sampling variance based on the idea of Rao-Blackwellization is also explored. Illustrative examples include identification of multi-component and spatially varying support motions in linear/nonlinear structures.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• 제25권3B호
• /
• pp.464-471
• /
• 2000

The number of quantization bits of the input signals $X_k$,$Y_k$ need to be optimally determined through the trade-off between the H/W complexity and the BER performance in Turbo codes applications. Also, an effective means to incorporate a channel reliability $L_c$ in the Log-MAP-based Turbo decoding is highly required. because it has a major effect on both the complexity and the performance. In this paper, a novel bit-shifting approach that substitutes for the multiplying is proposed so as to effectively incorporate. $L_c$ in Turbo decoding. The optimal number of quantization bits of $X_k$,$Y_k$ is investigated through Monte-Carlo simulations assuming that bit-shifting approach is adopted. In addition. The effects of an incorrect estimation of noise variance on the performance of Turbo codes is investigated. There is a confined range in which the effects of an incorrect estimation can be ignored.