• 대한전기학회:학술대회논문집
• /
• 대한전기학회 2004년도 학술대회 논문집 정보 및 제어부문
• /
• pp.343-345
• /
• 2004

We propose a new category of fuzzy set-based fuzzy inference systems based on data granulation related to fuzzy space division for each variables. Data granules are viewed as linked collections of objects(data, in particular) drawn together by the criteria of proximity, similarity, or functionality. Granulation of data with the aid of Hard C-Means(HCM) clustering algorithm help determine the initial parameters of fuzzy model such as the initial apexes of the membership functions and the initial values of polyminial functions being used in the premise and consequence part of the fuzzy rules. And the initial parameters are tuned effectively with the aid of the genetic algorithms(GAs) and the least square method. Numerical example is included to evaluate the performance of the proposed model.

• 한국대기환경학회지
• /
• 제23권6호
• /
• pp.653-665
• /
• 2007

Several numerical experiments were carried out to clarify the influence of satellite data assimilation with various spatial resolution on mesoscale meteorological wind and temperature field. Satellite data used in this study is QuikSCAT launched on ADEOS II. QuikSCAT data is reasonable and faithful sea wind data, which have been verified through many observational studies. And numerical model in the study is MM5 developed by NCAR. Difference of wind pattern with and without satellite data assimilation appeared clearly, especially wind speed dramatically reduced on East Sea, when satellite data assimilation worked. And sea breeze is stronger in numerical experiments with RDAPS and satellite data assimilation than that with CDAS and data assimilation. This caused the lower estimated surface temperature in CDAS used cases. Therefore the influence of satellite data assimilation acts differently according to initial data quality. And it is necessary to make attention careful to handle the initial data for numerical simulations.

• 대한조선학회논문집
• /
• 제43권3호
• /
• pp.362-374
• /
• 2006

Currently, all design information of a hull structure is being first defined on 2D drawings not 3D CAD model at the initial ship design stage and then transferred to following design stages through the 2D drawings. This is caused by the past design practice, limitation on time, and lack of hull structural CAD systems supporting the initial design stage. As a result, the following design tasks such as the process planning and scheduling are being manually performed using the 2D drawings. For solving this problem, a data structure supporting the initial design stage is proposed and a prototype system is developed based on the data structure. The applicability of the system is demonstrated by applying it to various examples. The results show that the system can be effectively used for generating the 3D CAD model of the hull structure at the initial design stage.

• 한국경영과학회지
• /
• 제33권3호
• /
• pp.93-105
• /
• 2008

Semi-supervised learning uses a small amount of labeled data to predict labels of unlabeled data as well as to improve clustering performance, whereas unsupervised learning analyzes only unlabeled data for clustering purpose. We propose a new clustering-based semi-supervised learning method by reflecting the initial predicted labels of unlabeled data on the objective function. The initial prediction should be done in terms of a discrete probability distribution through a classification method using labeled data. As a result, clusters are formed and labels of unlabeled data are predicted according to the Information of labeled data in the same cluster. We evaluate and compare the performance of the proposed method in terms of classification errors through numerical experiments with blinded labeled data.

• Journal of the Korean Data and Information Science Society
• /
• 제18권2호
• /
• pp.481-488
• /
• 2007

The k-means algorithm is well known for its efficiency in clustering large data sets. However, working only on numeric values prohibits it from being used to cluster real world data containing categorical values. The k-modes algorithm is to extend the k-means paradigm to categorical domains. The algorithm requires a pre-setting or random selection of initial points (modes) of the clusters. This paper improved the problem of k-modes algorithm, using the Max-Min method that is a kind of methods to decide initial values in k-means algorithm. we introduce new similarity measures to deal with using the categorical data for clustering. We show that the mushroom data sets and soybean data sets tested with the proposed algorithm has shown a good performance for the two aspects(accuracy, run time).

• 디지털융복합연구
• /
• 제16권12호
• /
• pp.217-229
• /
• 2018

이 연구에서는 미분양 아파트 단지의 세대별 계약 자료에 데이터 마이닝 기법인 의사결정나무, 신경망, 로지스틱 모형을 적용하여 세대별 초기계약을 예측하는 모형을 개발한다. 모형 개발에는 위례신도시 미분양 아파트 단지의 계약 자료가 이용되며, 이 자료는 훈련용 자료와 검정용 자료로 분할되어 분석에 투입된다. 훈련용 자료에서는 신경망, 의사결정나무, 로지스틱 모형 순으로 예측력이 뛰어났지만 검정용 자료에서는 로지스틱 모형이 가장 우수하게 나타났다. 이 같은 결과는 신경망이 훈련용 자료에 최적화된 모형으로 구축되면서 검정용 자료에 대한 적응성이 떨어져 나타난 결과로 판단된다. 의사결정나무와 로지스틱 모형을 병행 적용한 결과, 층수, 향, 세대 위치, 전기 및 발전기실의 소음, 청약자 거주지, 청약 종류가 초기계약에 영향을 주는 것으로 나타났다. 이는 두 가지 모형을 같이 사용하는 것이 초기계약 결정요인 발굴에 더 효과적이라는 것을 의미한다. 이 연구는 데이터 마이닝의 적용 범위를 주택 분양 예측까지 확장함으로써 융복합 분야 발전에 기여하고 있다.

• Journal of Microbiology and Biotechnology
• /
• 제13권3호
• /
• pp.332-337
• /
• 2003

Enzyme kinetics data play a vital role in the design of reactors and control of processes. In the present study, kinetic studies on pectinases were carried out. Partially purified polymethylgalacturonase (PMG) and polygalacturonase (PG) were the two pectinases studied. The plot of initial rate vs. initial substrate concentration did not follow the conventional Michaelis-Menten kinetics, but substrate inhibition was observed. For PMG, maximum rate was attained at an initial pectin concentration of 3 g/l, whereas maximum rate was attained when the initial substrate concentration of 2.5 g/l of polygalacturonic acid for PG I and PG II. The kinetic data were fitted to five different kinetic models to explain the substrate inhibition effect. Among the five models tested, the combined mechanism of protective diffusion limitation of both high and inhibitory substrate concentrations (semi-empirical model) explained the inhibition data with 96-99% confidence interval.

• 대한공업교육학회지
• /
• 제32권2호
• /
• pp.199-216
• /
• 2007

SCTP(stream control transmission protocol)는 데이터 전송을 위한 전송 계층 프로토콜로서, 많은 부분에서 TCP(transmission control protocol) 방식을 따른다. 하지만 멀티 호밍(multi-homing)과 멀티 스트리밍(multi-streaming)의 특징을 가짐으로 성능의 차이를 갖는다. 이 논문에서는 SCTP 혼잡제어 중에서 초기 슬로우 스타트 단계에 초점을 맞추어 데이터 전송을 분석하고, 대역폭, 지연시간 및 데이터 크기에 따른 SCTP와 TCP 평균 전송 시간을 측정하고 비교하였다. 아울러 SCTP와 TCP의 평균 전송시간에 영향을 미치는 요인인 초기 윈도우 크기를 데이터 크기에 따라 측정하였다. 실험을 위한 서버와 클라이언트 프로그램은 SCTP socket API를 이용하여 C 언어로 작성되었고, 전송 시간은 이더리얼 프로그램을 사용하여 측정되었다. 서버와 클라이언트 사이의 데이터 전송 방법은 라운드 로빈(round robin) 방법을 사용하였다. 실험 결과, SCTP는 초기 슬로우 스타트 단계에서 TCP 보다 평균 전송 시간에 있어 약 15% 정도 향상된 성능을 보였으며, 그 이유는 SCTP 초기 윈도우 크기가 TCP 보다 크기 때문으로 확인되었다.

• 한국과학교육학회지
• /
• 제20권2호
• /
• pp.288-296
• /
• 2000

본 연구에서는 변칙 사례에 대한 학생들의 반응 유형과 특성을 조사하였다. 학생들의 응답 분류 기준은 '변칙 사례의 타당성 인정', '변칙 사례와 초기 이론 사이의 불일치성 인정', 그리고 '초기 이론에 대한 확신의 변화' 등이었다. 분류 결과, 거부, 재해석, 배제, 판단 불가, 주변 이론의 변화, 신념의 일부 변화, 이론 변화 등 7가지 반응 유형을 얻었다. 초기 이론에 대한 무조건적인 신뢰나 실험 방법의 정확성에 대한 의심이 변칙 사례를 거부하는 주된 원인이었다. 학생들은 변칙 사례와 초기 이론에 관련된 실험 과정은 무시하고 실험 결과의 유사성에 더 주목했기 때문에 불일치성을 인정하지 않았다.

• Journal of Information Science Theory and Practice
• /
• 제2권2호
• /
• pp.6-15
• /
• 2014

Initial based disambiguation of author names is a common data pre-processing step in bibliometrics. It is widely accepted that this procedure can introduce errors into network data and any subsequent analytical results. What is not sufficiently understood is the precise impact of this step on the data and findings. We present an empirical answer to this question by comparing the impact of two commonly used initial based disambiguation methods against a reasonable proxy for ground truth data. We use DBLP, a database covering major journals and conferences in computer science and information science, as a source. We find that initial based disambiguation induces strong distortions in network metrics on the graph and node level: Authors become embedded in ties for which there is no empirical support, thus increasing their sphere of influence and diversity of involvement. Consequently, networks generated with initial-based disambiguation are more coherent and interconnected than the actual underlying networks, and individual authors appear to be more productive and more strongly embedded than they actually are.