• The Korean Journal of Applied Statistics
• /
• v.7 no.2
• /
• pp.225-237
• /
• 1994

A generalized likelihood ratio test is developed to detect an outlier associated with monitoring nuclear proliferation. While the classical outlier detection methods consider continuous variables only, our approach allows both continuous and discrete variables or a mixture of continuous and discrete variables to be used. In addition, our method is free of the normality assumption, which is the key assumption in most of the classical methods. The proposed test is constructed by applying the bootstrap to a generalized likelihood ratio. We investigate the performance of the test by studying the power with simulations.

• Proceedings of the Korean Information Science Society Conference
• /
• 2003.04a
• /
• pp.800-802
• /
• 2003

이 논문에서는 웹 기반의 하이퍼미디어 교육시스템에서 이산 확률 분포 함수와 사용자 프로파일 기반의 동적 적응 모델을 제안하였다. 이 모델은 응용 영역을 동적 적응 객체의 가중치 방향성 그래프로 표현하며 사용자 행위를 이산 확률 함수를 동적으로 구축하는 접근 방식을 이용하여 모델링 한다. 제안한 확률적 해석은 웹 미디어 구조에서 사용자의 탐색 행위를 추적하여 사용자 행위에 대한 잠재적 속성을 나타내는데 사용될 수 있다. 이러한 접근 방식은 사용자에게 가장 알맞은 프로파일을 동적으로 할당할 수 있다.

• Communications for Statistical Applications and Methods
• /
• v.17 no.2
• /
• pp.153-164
• /
• 2010

This study is about the dynamic graphics which can be used for the exploration of the characteristics of data comprising discrete and continuous variables. Simultaneously using line mosaic plot for the relation of discrete variables and box plot together with scatter plot for the relation of continuous variables, we have applied dynamic methods among these plots to demonstrate that the structure and characteristics of the multivariate data could be easily analyzed.

• Journal of the Korean Society for Aeronautical & Space Sciences
• /
• v.33 no.9
• /
• pp.56-65
• /
• 2005

A dynamic analysis method that freezes a time domain by discretization and solves the spatial propagation equation has a unique feature that provides a degree of freedom on spatial domain compared with the space discretization or space-time discretization finite element method. Using this feature, the time finite element analysis can be effectively applied to optimize the spatial characteristics of distributed type actuators. In this research, the time domain finite element method was used to discretize the model. A state variable vector was used in the discretization to include arbitrary initial conditions. A performance index was proposed on spatial domain to consider both potential and vibrational energy, so that the resulting shape of the distributed actuator was optimized for dynamic control of the structure. It is assumed that the structure satisfies the final rest condition using the realizable control scheme although the initial disturbance can affect the system response. Both equations on states and costates were derived based on the selected performance index and structural model. Ricatti matrix differential equations on state and costate variables were derived by the reconfiguration of the sub-matrices and application of time/space boundary conditions, and finally optimal actuator distribution was obtained. Numerical simulation results validated the proposed actuator shape optimization scheme.

• Journal of Korea Society of Industrial Information Systems
• /
• v.12 no.1
• /
• pp.108-117
• /
• 2007

For a PERT network the sensitivity analyses are valuable tools for effective management and control of a project. In a previous article, a method was developed for the sensitivity analysis of the expected project completion time (${\mu}_{T}$) with respect to the expected duration of an activity (${\mu}_{i}$) under the assumption that the durations of activities are independent and normally distributed. In the present article, a method is presented for the sensitivity analysis of ${\mu}_{T}$ with respect to ${\mu}_{i}$ when the durations of activities are independent and follow discrete distributions, which utilizes the previous sensitivity analysis method with modifications. The accuracy of the proposed method is assessed using various example PERT networks, and the computational results indicate that the accuracy is comparable to that of direct Monte Carlo simulation.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.13 no.4
• /
• pp.391-396
• /
• 2003

This paper proposes a new approach that converts continuous-valued attributes to categorical-valued ones considering the distribution of target attributes(classes). In this approach, It can be possible to get optimal interval boundaries by considering the distribution of data itself without any requirements of parameters. For each attributes, the distribution of target attributes is projected to one-dimensional space. And this space is clustered according to the criteria like as the density value of each target attributes and the amount of overlapped areas among each density values of target attributes. Clusters which are made in this ways are based on the probabilities that can predict a target attribute of instances. Therefore it has an interval boundaries that minimize a loss of information of original data. An improved performance of proposed discretization method can be validated using C4.5 algorithm and UCI Machine Learning Data Repository data sets.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.20 no.2
• /
• pp.165-172
• /
• 2010

In clinical data minig, choosing the optimal subset of features is such important, not only to reduce the computational complexity but also to improve the usefulness of the model constructed from the given data. Moreover the threshold values (i.e., cut-off points) of selected features are used in a clinical decision criteria of experts for differential diagnosis of diseases. In this paper, we propose a fuzzy discretization approach, which is evaluated by measuring the degree of separation of redundant attribute values in overlapping region, based on spatial distribution of data with continuous attributes. The weighted average of the redundant attribute values is then used to determine the threshold value for each feature and rough set theory is utilized to select a subset of relevant features from the overall features. To verify the validity of the proposed method, we compared experimental results, which applied to classification problem using 668 patients with a chief complaint of dyspnea, based on three discretization methods (i.e., equal-width, equal-frequency, and entropy-based) and proposed discretization method. From the experimental results, we confirm that the discretization methods with fuzzy partition give better results in two evaluation measures, average classification accuracy and G-mean, than those with hard partition.

• The Journal of Korea Institute of Information, Electronics, and Communication Technology
• /
• v.11 no.2
• /
• pp.169-174
• /
• 2018

In general, the methods of the analysis of variance(ANOVA) for the continuous data and the chi-square test for the discrete data are used for statistical analysis of the effect and the association. In multidimensional data, analysis of hierarchical structure is required and statistical linear model is adopted. The structure of the linear model requires the normality of the data. A multidimensional categorical data analysis methods are used for causal relations, interactions, and correlation analysis. In this paper, Bayesian network model using probability distribution is proposed to reduce analysis procedure and analyze interactions and causal relationships in categorical data analysis.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2011.05a
• /
• pp.326-330
• /
• 2011

댐의 최적운영을 위해서는 강우량, 유량, 토양수분량, 증발산량과 같은 수문자료는 필수적이다. 이중 강우량과 유량자료는 치수 중심의 댐 운영에 가장 중요하게 이용되며, 국가 수자원계획, 이수 및 환경 계획 등에도 다목적으로 활용된다. 강우량은 면적 강우량을 대표할 수 있는 위치에서 관측되어야 점 강우량을 면적 강우량으로 환산하는데서 발생되는 오차를 최소화할 수 있다. 이는 실제 발생되는 연속형 강우량과 강우관측소에서 관측되는 이산형 강우량의 차가 최소화될 때 가능한 일이다. 최근 강우 특성은 급 점진적으로 변화하고 있다. 과거에 비해 매우 시공간적으로 불규칙해졌으며, 특히 짧은 지속시간 동안 많은 양의 강우가 집중되고 있다. 이와 같은 강우 특성 변화는 강우관측망에 반드시 반영되어야 한다. 강우 특성을 반영하여 댐을 효율적으로 운영하기 위해서는 기존 관측망에 대한 재평가가 선행되어야 하며, 재평가된 결과를 토대로 관측망을 개선해야 한다. 이에 따라 본 연구에서는 최근 10개년(기상청)의 강우자료를 Kriging method로 공간 분포시켜 연속형 강우량과 강우관측소에서 관측되는 이산형 강우량의 차가 최소화될 수 있는 강우관측망을 구축하였다. 강우관측망을 구축한 결과, 최소 72개소의 강우관측소가 필요하였다. 기관별로는 한수원(주) 29개소(화천댐 유역, 신설 2개소 포함), 국토해양부 18개소, 한국수자원공사 4개소, 기상청(유인 및 무인) 21개소로 구축되었다. 본 연구에서 설계한 강우관측망은 대략 평균 $100km^2$의 밀도로 구축되었으며, 팔당댐 유역에서 가장 크게 개선되었다.

• Korean Journal of Soil Science and Fertilizer
• /
• v.43 no.6
• /
• pp.1035-1039
• /
• 2010

For 65 national forest areas in 1993 to 2008, the ambient sulfur dioxide ($SO_2$) concentrations were measured monthly using passive samplers and compared to the those of urban areas in order to investigate the characteristics of temporal and spatial distribution. In the forest areas, annual average concentration of sulfur dioxide gradually decreased from the beginning year of monitoring to 1997 and then had no significant change, such as the annual trend in urban areas monitored by Ministry of Environment. For the monitoring term, average concentration of sulfur dioxide in the forest areas was 5.6ppb, which was lower than the 10.1ppb in the urban areas and the EC ecological standard level (7.6 ppb). Seasonally, both in forest areas and urban areas the monthly average concentrations were much higher in winter and spring due to much more heating fuel consumption, and lowest in summer. Regional comparison to other regions of Gyeongbuk and Gyeonggi province showed that the concentration of sulfur dioxide was the highest during year. A significant positive correlation between sulfur oxides' emissions and sulfur dioxide concentration by province was observed, reflecting that the size and proximity of sources of atmospheric sulfur oxides could be important factors in determination.