http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
An Website Evaluation Methodology
권세혁 한국데이터정보과학회 2008 한국데이터정보과학회지 Vol.19 No.1
Websites are critical and necessary marketing tools to companies for providing the business information to customers and understanding customers' needs and huge information warehouses to users for searching what they need to know. Hence, the size and number of websites even in a company rapidly increase, which results in the redundancy and increasing similarity of websites. For the cost and time effectiveness to companies and customers, the websites should be evaluated and ranked. In general, websites have been quantitatively evaluated as follows: (1)the subjective and qualitative assessment (for example, Likert scale) of experts on the designated evaluation items is collected and (2)the arithmetic and/or weighted (by the subjective choice) mean of those assessments is computed. In this paper, AHP methodology is suggested to obtain weights of evaluation items and an empirical result is given to show how to use it.
Obtaining bootstrap data for the joint distribution of bivariate survival times
권세혁 한국데이터정보과학회 2009 한국데이터정보과학회지 Vol.20 No.5
The bivariate data in clinical research elds often has two types of failure times, which are mark variable for the rst failure time and the nal failure time. This paper showed how to generate bootstrap data to get Bayesian estimation for the joint distribution of bivariate survival times. The observed data was generated by Frank's family and the fake date is simulated with the Gamma prior of survival time. The bootstrap data was obtained by combining the mimic data with the observed data and the simulated fake data from the observed data.
대전력계통(大電力系統) 운용시 수수(授受) 에너지 오차의 차대변(借貸邊) 계정에 관한 연구
권세혁 고려대학교 공학기술연구소 1987 고려대학교 생산기술연구소 생기연논문집 Vol.23 No.1
The inadvertent interchange energy of any control areas on the interconnected power systems can be decomposed into primary component and secondary component. If some control areas do not want to take the corrective control actions because of increasing regulating coats and increasing disparity in incremental costs during the course of a day, a debit/credit system can be used to determine the associated regulation costs. Debit/Credit system suggested by Nathan Cohn has been studied and a new Dibit/Credit system is suggested to overcome the complexities in computations. It is believed that the computations of the debit and credit of controls areas becomes quite simple.
무한모선(無限母線)에 연결된 동기발전기(同期發電機)의 모델링방법에 대한 비교연구
권세혁,김덕영 고려대학교 공학기술연구소 1989 고려대학교 생산기술연구소 생기연논문집 Vol.25 No.1
The dominant eigenvalue pairs of a synchronous generator connected to an infinite bus for three kinds of modelling have been compared. An efficient calculating procedure of the elements of A matrix of a set of differential equations have been suggested for a full model. The submatrices of A matrix which depends on the machine and transmission parameters sie computed just once and the submatrices which depend on the system parameters and the initial condition are calculated for each initial condition. One-axis model wan derived from a full model and a simplified linear model is a special case of one-axis model which neglects the term related to the speed deviation and the stator winding resistance. A full model, an one-axis model, and a simplified linear model for the same machine for the same initial condition are compared. One-axis model has the loci of the dominant eigenvalues nearest to the imaginary axis.
머신러닝 기법을 활용한 대용량 시계열 데이터 이상 시점탐지 방법론 : 발전기 부품신호 사례 중심
권세혁 한국산업경영시스템학회 2020 한국산업경영시스템학회지 Vol.43 No.2
Anomaly detection of Machine Learning such as PCA anomaly detection and CNN image classification has been focused on cross-sectional data. In this paper, two approaches has been suggested to apply ML techniques for identifying the failure time of big time series data. PCA anomaly detection to identify time rows as normal or abnormal was suggested by converting subjects identification problem to time domain. CNN image classification was suggested to identify the failure time by re-structuring of time series data, which computed the correlation matrix of one minute data and converted to tiff image format. Also, LASSO, one of feature selection methods, was applied to select the most affecting variables which could identify the failure status. For the empirical study, time series data was collected in seconds from a power generator of 214 components for 25 minutes including 20 minutes before the failure time. The failure time was predicted and detected 9 minutes 17 seconds before the failure time by PCA anomaly detection, but was not detected by the combination of LASSO and PCA because the target variable was binary variable which was assigned on the base of the failure time. CNN image classification with the train data of 10 normal status image and 5 failure status images detected just one minute before.
권세혁,이요상,Kwon, Se-Hyug,Lee, Yo-Sang 한국통계학회 2010 Communications for statistical applications and me Vol.17 No.1
환경가치가 높아짐에 따라 하천 수질에 대한 관심의 증대로 수질측정망 연구가 최근 활발히 진행되고 있으나 입지환경의 지리적 특성이나 유입량, 유출량, 유량, 유속과 같은 하천 특성 중심 연구이다. 본 연구에서는 상대적으로 연구가 미미한 관측지점의 수질 유사성을 측정하는 방법으로 수질의 시계열 패턴을 고려할 수 있는 상관계수행렬 방법을 제안하고 기존의 주성분점수를 이용한 방법과 비교하였다. 용담댐에서 2년간 조사된 수질관련 데이터를 이용하여 두 방법에 대한 실증분석을 실시하여 관측지점의 지리적 특성에 의해 분류된 결과와 본 연구에서 제안된 방법에 의해 관측지점 유사성을 측정하여 얻은 군집결과가 더 일치함을 보였다. As the value of environment is increasing, the water quality has been a matter of interest to the nation and people. Research on water quality has been widely studied, but focused on geographical characteristic and river characteristics like inflow, outflow, quantity and speed of water. In this paper, two approaches to measure the similarity of sampling sites by using water quality data are discussed and compared with two-years empirical data of Yongdam-Dam. The existing method has calculated their similarities with principal component scores. The proposed approach in this paper use correlation matrix of water quality related variables and MDS for measuring the similarity, which is shown to be better in the sense of being clustering which is identical to geographical clustering since it can consider the time series pattern of water quality.