• Journal of the Institute of Electronics Engineers of Korea CI
• /
• v.44 no.1
• /
• pp.26-32
• /
• 2007

In sequential speaker classification, the lack of prior information about the speakers poses a challenge for model initialization. To address the challenge, a predetermined generic model set, called Sample Speaker Models, was previously proposed. This approach can be useful for accurate speaker modeling without requiring initial speaker data. However, an optimal method for sampling the models from a generic model pool is still required. To solve this problem, the Speaker Quantization method, motivated by vector quantization, is proposed. Experimental results showed that the new approach outperformed the random sampling approach with 25% relative improvement in error rate on switchboard telephone conversations.

• Proceedings of the KSRS Conference
• /
• v.2
• /
• pp.816-819
• /
• 2006

One of the most difficult parameters to measure in the sea is current speed and direction. Recently, efforts are being made to estimate the ocean current vectors by utilizing sequential satellite imageries. In this study, we attempted to estimated sea surface current vector (sscv) by using satellite ocean color imageries of SeaWifs around the Korean Peninsula. This ocean color image data has 1-day sampling interval and spatial resolution of 1x1 km. Maximum cross-correlation method is employed which is aimed to detect similar patterns between sequential images. The estimated current vectors are compared to the surface geostrophic current vectors obtained from altimeter of sea level height data. In utilizing the color imagery data, some limitations and drawbacks exist so that in warm water region where phytoplankton concentration is relatively lower than in cold water region, estimation of sscv is poor and unreliable. On the other hand, two current vector fields agree reasonably well in the Korean South Sea region where high concentration of chlorophyll-a and weak tide is observed. In the future, with ocean color images of shorter sampling interval by COMS satellite, the algorithm and methodology developed in the study would be useful in providing the information for the ocean current around Korean Peninsula.

• Korean Journal of Computational Design and Engineering
• /
• v.17 no.5
• /
• pp.375-386
• /
• 2012

Most engineering design problems require analyses or simulations to evaluate objective functions. However, a single simulation can take many hours or even days to finish for many real world problems. As a result, design optimization becomes impossible since they require hundreds or thousands of simulation evaluations. The surrogate-based optimization (SBO) strategy became a remedy for such computationally expensive analyses and simulations. A surrogate-based optimization strategy has been developed in this study in order to improve global optimization performance. The strategy is a heuristic algorithm and it exploits not only multiple surrogates, but also multiple optimizers. Multiple optimizations of multiple surrogate models yield multiple candidate design points of optima. During the sequential sampling process, the algorithm ranks candidate design points, selects the points as many as specified, and builds the improved surrogate model. Various mathematical functions with different numbers of design variables are chosen to compare the proposed method with the other most recent algorithm, MSEGO. The proposed method shows superior performance to the other method.

• Smart Structures and Systems
• /
• v.21 no.5
• /
• pp.601-609
• /
• 2018

The estimated probabilistic model of wind data based on the conventional approach may have high discrepancy compared with the true distribution because of the uncertainty caused by the instrument error and limited monitoring data. A sequential quadratic programming (SQP) algorithm-based finite mixture modeling method has been developed in the companion paper and is conducted to formulate the joint probability density function (PDF) of wind speed and direction using the wind monitoring data of the investigated bridge. The established bivariate model of wind speed and direction only represents the features of available wind monitoring data. To characterize the stochastic properties of the wind parameters with the subsequent wind monitoring data, in this study, Bayesian inference approach considering the uncertainty is proposed to update the wind parameters in the bivariate probabilistic model. The slice sampling algorithm of Markov chain Monte Carlo (MCMC) method is applied to establish the multi-dimensional and complex posterior distribution which is analytically intractable. The numerical simulation examples for univariate and bivariate models are carried out to verify the effectiveness of the proposed method. In addition, the proposed Bayesian inference approach is used to update and optimize the parameters in the bivariate model using the wind monitoring data from the investigated bridge. The results indicate that the proposed Bayesian inference approach is feasible and can be employed to predict the bivariate distribution of wind speed and direction with limited monitoring data.

• The Journal of Korean Institute of Electromagnetic Engineering and Science
• /
• v.26 no.6
• /
• pp.546-554
• /
• 2015

Ultra-wideband(UWB) radar is widely used in many penetration radar applications, such as ground-penetrating radar and foliage-penetrating radar, because it has many advantages in detecting concealed objects. One type of UWB radar system is random noise radar, which many be robust to jamming environment. However conventional random noise radar requires high-speed analog-to-digital convertor(ADC) for matched filtering. In this thesis, a pseudo-random noise radar system that maintains anti-jamming characteristics but does not require high-speed ADC is researched. and The UWB system is implemented in a low frequency system, and its performance has been demonstrated by experiment, which proves the concept of the proposed pseudo-random noise radar system.

• Korean journal of applied entomology
• /
• v.42 no.1
• /
• pp.29-34
• /
• 2003

Dispersion pattern for adult citrus red mite (CRM), Panonychus citri (McGregor) using by Taylor's power law (TPL) and Iwao's patchiness regression (IPR) was determined to develop a monitoring method on citrus orchards, on Jeju, in Autumn season, during 2001 and 2002.CRM population was sampled by collecting leaves and fruits. The relationships of CRM adults between leaf and fruit were analyzed by different season. The regression equation for CRM adults between leaf (X) and fruit (Y) was ln(Y+1) : 1.029 ln(X+1) ( $r^2$ : 0.80). The density of CRM was higher on fruit than on leaf according to fruit maturing level. TPL provided better description of mean-variance relation-ship for the dispersion indices compared to IPR. Slopes and intercepts of TPL from leaf and fruit samples did not differ between sample units and surveyed years. Fixed-precision levels (D) of a sequential sampling plan were developed using Taylor's power law parameters generated from adults of CRM in leaf sample. Sequential sampling plans for adults of CRM were developed for decision making CRM population level based on the different action threshold levels (2.0,2.5 and 3.0 mites per leaf) with 0.25 precision. The maximum number of trees and required number of trees sampled on fixed sample size plan on 2.0,2.5 and 3.0 thresholds with 0.25 precision level were 19, 16 and 15 and their critical values T$_{critical}$ at were 554,609 and 659, respectively. were 554,609 and 659, respectively.

• Journal of Navigation and Port Research
• /
• v.30 no.1 s.107
• /
• pp.105-111
• /
• 2006

A LMTT (Linear Motor-based Transfer Technology) is a horizontal transfer system for the yard automation This system is based on PMLSM (Permanent Magnetic Linear Synchronous Motor) that consists of stator modules on the rail and shuttle car. In this research, the kriging interpolation method using sequential sampling is utilized to find the optimum design of a mover in LMTT. The design variables are considered as the transverse, longitudinal and wheel beam's thicknesses. The objective function is set up as weight, while the constant functions are set up as the stresses generated by four loading conditions. The optimum results obtained by the suggested method are compared with those determined by the GENESIS.

• Journal of Communications and Networks
• /
• v.14 no.2
• /
• pp.188-194
• /
• 2012

Cognitive networks (CNs) are capable of enabling dynamic spectrum allocation, and thus constitute a promising technology for future wireless communication. Whereas, the implementation of CN will lead to the requirement of an increased energy-arrival rate, which is a significant parameter in energy harvesting design of a cognitive user (CU) device. A well-designed spectrum-sensing scheme will lower the energy-arrival rate that is required and enable CNs to self-sustain, which will also help alleviate global warming. In this paper, spectrum sensing in a multi-user cognitive ad hoc network with a wide-band spectrum is considered. Based on the prospective spectrum sensing, we classify CN operation into two modes: Distributed and centralized. In a distributed network, each CU conducts spectrum sensing for its own data transmission, while in a centralized network, there is only one cognitive cluster header which performs spectrum sensing and broadcasts its sensing results to other CUs. Thus, a wide-band spectrum that is divided into multiple sub-channels can be sensed simultaneously in a distributed manner or sequentially in a centralized manner. We consider the energy consumption for spectrum sensing only of an analog-to-digital convertor (ADC). By formulating energy consumption for spectrum sensing in terms of the sub-channel sampling rate and whole-band sensing time, the sampling rate and whole-band sensing time that are optimal for minimizing the total energy consumption within sensing reliability constraints are obtained. A power dissipation model of an ADC, which plays an important role in formulating the energy efficiency problem, is presented. Using AD9051 as an ADC example, our numerical results show that the optimal sensing parameters will achieve a reduction in the energy-arrival rate of up to 97.7% and 50% in a distributed and a centralized network, respectively, when comparing the optimal and worst-case energy consumption for given system settings.

• Journal of the Korean Home Economics Association
• /
• v.46 no.9
• /
• pp.125-135
• /
• 2008

The purpose of this study was to examine the effects of young children's age and information processing style in understanding spatial coordinates. For sampling the subjects of this study, Korean version K-ABC Intelligence Test(Moon, Soo-Back, 1997)was conducted with 165 children aged 5-6 who were attending I and G kindergarten in D city. From this pool 30 children who possessed sequential processing style and 30 children who possessed simultaneous processing style were sampled. In order to analyze the understanding of spatial coordinates, a test tool was formulated according to methodology of Blades & Spencer(1989) which was modified. Acquired data was subjected to descriptive and comparative statistical analysis. The following conclusions were arrived at: Firstly, there was significant difference between 5-year-olds and 6-year-olds in understanding spatial coordinates. The 6-year-old group got statistically higher grades than the 5-year-old group in locating a point on the coordinate plane and reading the coordinate numbers. Secondly, there was significant difference between children's information processing style in understanding spatial coordinate. Children with high simultaneous-low sequential processing showed higher performance in locating a point on the coordinate plane and reading coordinate numbers than children with high sequential-low simultaneous processing. Thirdly, after verifying statistical significance of interactivity between young children's age and children's processing strength, there was significant interactive effects in both tasks.

• Communications for Statistical Applications and Methods
• /
• v.23 no.2
• /
• pp.131-146
• /
• 2016

In research on behavioral studies, significant attention has been paid to the stage-sequential process for longitudinal data. Latent class profile analysis (LCPA) is an useful method to study sequential patterns of the behavioral development by the two-step identification process: identifying a small number of latent classes at each measurement occasion and two or more homogeneous subgroups in which individuals exhibit a similar sequence of latent class membership over time. Maximum likelihood (ML) estimates for LCPA are easily obtained by expectation-maximization (EM) algorithm, and Bayesian inference can be implemented via Markov chain Monte Carlo (MCMC). However, unusual properties in the likelihood of LCPA can cause difficulties in ML and Bayesian inference as well as estimation in small samples. This article describes and addresses erratic problems that involve conventional ML and Bayesian estimates for LCPA with small samples. We argue that these problems can be alleviated with a small amount of prior input. This study evaluates the performance of likelihood and MCMC-based estimates with the proposed prior in drawing inference over repeated sampling. Our simulation shows that estimates from the proposed methods perform better than those from the conventional ML and Bayesian method.