• Structural Engineering and Mechanics
• /
• v.25 no.3
• /
• pp.331-345
• /
• 2007

Uncertainties enter a complex analysis from a variety of sources: variability, lack of data, human errors, model simplification and lack of understanding of the underlying physics. However, for many important engineering applications insufficient data are available to justify the choice of a particular probability density function (PDF). Sometimes the only data available are in the form of interval estimates which represent, often conflicting, expert opinion. In this paper we demonstrate that Bayesian estimation techniques can successfully be used in applications where only vague interval measurements are available. The proposed approach is intended to fit within a probabilistic framework, which is established and widely accepted. To circumvent the problem of selecting a specific PDF when only little or vague data are available, a hierarchical model of a continuous family of PDF's is used. The classical Bayesian estimation methods are expanded to make use of imprecise interval data. Each of the expert opinions (interval data) are interpreted as random interval samples of a parent PDF. Consequently, a partial conflict between experts is automatically accounted for through the likelihood function.

• Journal of the Korean Data and Information Science Society
• /
• v.7 no.2
• /
• pp.247-256
• /
• 1996

In this paper, we applied bayesian hierarchical model to analyze the field mice example introduced by Demster et al.(1981). For this example, we use Gibbs sampler method to provide the posterior mean and compared it with LSE(Least Square Estimator) and MLR(Maximum Likelihood estimator with Random effect) via the EM algorithm.

• ALGAE
• /
• v.17 no.1
• /
• pp.33-39
• /
• 2002

Mechanisms of competition between two canopy algae and an understory alga were investigated by a field manipulative experiment using artificial thalli. The study was carried out in the upper intertidal zone at Nudibranch Point in Vancouver Island, British Columbia, Canada, where two fucoids, Fucus gardneri and Pelvetiopsis limitata, and a turf red alga, Mazzaella cornucopiae, were dominant in the algal community. The experiment was designed to test three hypotheses, shading, whiplash, and allelopathy, imposed by potential fucoid effects on M. cornucopiae. Only the shading effect was significant, indicating that adult fucoid thalli reduced. M. cornucopiae biomass underneath the fucoids. Results indicated that reversal of competitive dominance existed between F. gardneri and M. cornucopiae depending on the life history stage of the competitors. By including the turf alga's effects on the fucoids, the well-balanced and non-hierarchical interaction networks among the major macroalgae support the high likelihood of species coexistence in the community.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2007.05a
• /
• pp.1081-1083
• /
• 2007

The study in river basin has been performed for the identify water quality impaired stream segments, to create a priority ranking of those segments, and to calculate the heavy metal ion distribution for each impaired segment based on chemical and physical water quality standards. Two methods for modeling the potential area-specific heavy metal distribution are pursued in this study. First, a novel approach focuses on distance. Heavy metal distribution can be associated with a particular small geographic area. Based on the derived estimates an distribution map can be generated. Second, the approach is used the near watershed by means of kriging interpolation algorithm. These approaches provide an alternative distribution mapping of the area. The exposure estimates from both of these modeling methods are then compared with other environmental monitoring data. A GIS-based model will be used to mimic the hierarchical stream structure and processes found in natural watershed. Specifically, the relationship between landscape variables and reach scale habitat conditions most influential found in the Abandoned mine will be explored.

• Journal of Communications and Networks
• /
• v.9 no.2
• /
• pp.167-176
• /
• 2007

In multimedia communication systems, efficient transmission system design should incorporate the use of matching unequal error protection (UEP), since source coders exhibit unequal bit error sensitivity. In this paper, we present UEP schemes which exploit differences in bit error protection levels in orthogonal frequency division multiplexing (OFDM) systems over frequency selective fading channels. We introduce an UEP scheme which improves the link performance with multiple transmit and receive antennas. Especially, we propose a new receiver structure based on two stage Maximum Likelihood detection (MLD) schemes which can approach the performance of a full search MLD receiver with much reduced computational complexity. In the performance analysis, we derive a generalized pairwise error probability expression for the proposed UEP schemes. Simulation results show that the proposed schemes achieve a significant performance gain over the conventional equal error protection (EEP) scheme.

• Korean Journal of Remote Sensing
• /
• v.20 no.5
• /
• pp.353-359
• /
• 2004

This study proposes an approach to unsupervisedly estimate the number of classes and the parameters of defining the classes in order to train the classifier. In the proposed method, the image is segmented using a spatial region growing based on hierarchical clustering, and fuzzy training is then employed to find the sample classes that well represent the ground truth. For cluster validation, this approach iteratively estimates the class-parameters in the fuzzy training for the sample classes and continuously computes the log-likelihood ratio of two consecutive class-numbers. The maximum ratio rule is applied to determine the optimal number of classes. The experimental results show that the new scheme proposed in this study could be used to select the regions with different characteristics existed on the scene of observed image as an alternative of field survey that is so expensive.

• The Korean Journal of Applied Statistics
• /
• v.28 no.5
• /
• pp.965-976
• /
• 2015

Determining relevant variables for a regression model is important in regression analysis. Recently, a variable selection methods using a penalized likelihood with various penalty functions (e.g. LASSO and SCAD) have been widely studied in simple statistical models such as linear models and generalized linear models. The advantage of these methods is that they select important variables and estimate regression coefficients, simultaneously; therefore, they delete insignificant variables by estimating their coefficients as zero. We study how to select proper variables based on penalized hierarchical likelihood (HL) in semi-parametric frailty models that allow three penalty functions, LASSO, SCAD and HL. For the variable selection we develop a new function in the "frailtyHL" R package. Our methods are illustrated with breast cancer survival data from the Medical Center at Chonnam National University in Korea. We compare the results from three variable-selection methods and discuss advantages and disadvantages.

• Journal of Korea Water Resources Association
• /
• v.46 no.1
• /
• pp.13-24
• /
• 2013

The main objective of this study was to develop a new regional frequency analysis model based on hierarchical Bayesian model that allows us to better estimate and quantify model parameters as well as their associated uncertainties. A Monte-carlo experiment procedure has been set up to verify the proposed regional frequency analysis. It was found that the proposed hierarchical Bayesian model based regional frequency analysis outperformed the existing L-moment based regional frequency analysis in terms of reducing biases associated with the model parameters. Especially, the bias is remarkably decreased with increasing return period. The proposed model was applied to six weather stations in Jeollabuk-do, and compared with the existing L-moment approach. This study also provided shrinkage process of the model parameters that is a typical behavior in hierarchical Bayes models. The results of case study show that the proposed model has the potential to obtain reliable estimates of the parameters and quantitatively provide their uncertainties.

• Korean Journal of Remote Sensing
• /
• v.27 no.4
• /
• pp.435-444
• /
• 2011

The land cover map derived from spectral features of high resolution optical images has low spectral resolution and heterogeneity in the same land cover class. For this reason, despite the same land cover class, the land cover can be classified into various land cover classes especially in vegetation area. In order to overcome these problems, detailed vegetation classification is applied to optical satellite image and SAR(Synthetic Aperture Radar) integrated data in vegetation area which is the result of pre-classification from optical image. The pre-classification and vegetation classification were performed with MLC(Maximum Likelihood Classification) method. The hierarchical land cover classification was proposed from fusion of detailed vegetation classes and non-vegetation classes of pre-classification. We can verify the facts that the proposed method has higher accuracy than not only general SAR data and GLCM(Gray Level Co-occurrence Matrix) texture integrated methods but also hierarchical GLCM integrated method. Especially the proposed method has high accuracy with respect to both vegetation and non-vegetation classification.

• Korean Journal of Health Education and Promotion
• /
• v.26 no.2
• /
• pp.1-13
• /
• 2009

Objectives: This study identified how personal characteristics, family environment, governmental policy for the prevention and cessation of smoking might influence on adolescent smoking. Methods: This study used data from the 2006 Korea Youth Risk Behavior Web-based Survey of 71,404 middle school and high school students, giving a response rate of 90.9%. We selected 61,508 adolescents subjects of the final analysis without missing data on independent variables and dependent variables which are used in this study. This study used $\chi^2$ tests and logistic regression models. Variables were added to the regression model in three groups using a hierarchical approach.Results: Adolescents were significantly more likely to become current smokers if they were boys, were in a higher grade, and had lower academic achievement. Adolescents experiencing stress and depression were associated with increased risk of current smoking. Adolescents with single parents or students of non-living with parents comparing with students of living with parents showed the high possibility of smoking. Lower father's education was associated with increased likelihood of current smoking. Adolescents who were exposed to smoking at home were more likely to smoke. Adolescents without contacting with the antismoking media campaign was associated with increased likelihood of current smoking. Conclusion: Promoting antismoking media campaigns targeted at adolescent is required, and the smoking prevention education which are proper for subjects are required. Proper plans which could decrease the exposure of secondhand smoking should be established.