• Journal of the Korean Institute of Gas
• /
• v.22 no.2
• /
• pp.52-58
• /
• 2018

The flash point is one of the most important properties to characterize fire and explosion hazard of flammable liquid mixture. In this paper, the flash points of ternary liquid mixture, n-nonane+n-decane+n-tridecane system, were measured using Seta flash closed cup tester. The measured values were compared with the calculated values using Raoult's law and multiple regression analysis. The absolute average errors(AAE) of the results calculated by Raoult's law is $0.6^{\circ}C$. The absolute average errors of the results calculated by multiple regression analysis is $0.4^{\circ}C$. As can be seen from AAE, the calculated values based on multiple regresstion analysis were found to be better than those based on Raoult's law.

• Journal of Satellite, Information and Communications
• /
• v.11 no.4
• /
• pp.94-101
• /
• 2016

Satellite data for remote sensing technology has limitations, especially with visible range sensor, cloud and/or other environmental factors cause missing data. In this study, using land surface temperature data from the MODerate resolution Imaging Spectro-radiometer(MODIS), we developed retrieving methods for satellite missing data and developed three methods; mean bias, regression analysis and local variation method. These methods used the previous day data as reference data. In order to validate these methods, we selected a specific measurement ratio using artificial missing data from 2014 to 2015. The local variation method showed low accuracy with root mean square error(RMSE) more than 2 K in some cases, and the regression analysis method showed reliable results in most cases with small RMSE values, 1.13 K, approximately. RMSE with the mean bias method was similar to RMSE with the regression analysis method, 1.32 K, approximately.

• Asia-Pacific Journal of Business Venturing and Entrepreneurship
• /
• v.7 no.2
• /
• pp.101-111
• /
• 2012

In the ALC(Autoclaved lightweight concrete) manufacturing process, if the pre-cured semi-cake is removed after proper time is passed, it will be hard to retain the moisture and be easily cracked. Therefore, in this research, we took the research by multiple regression analysis to find relationship between variables for the prediction the hardness that is the control standard of the removal time. We study the relationship between Independent variables such as the V/T(Vibration Time), V/T movement, expansion height, curing time, placing temperature, Rising and C/S ratio and the Dependent variables, the hardness by multiple regression analysis. In this study, first, we calculated regression equation by the regression analysis, then we tried phased regression analysis, best subset regression analysis and residual analysis. At last, we could verify curing time, placing temperature, Rising and C/S ratio influence to the hardness by the estimated regression equation.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2023.05a
• /
• pp.522-522
• /
• 2023

지하수 시스템의 방출은 저지대 강에서 건조기에 흐르는 하천 유지유량의 원천이 된다. 수자원 분야에서 분포형 모형이 도입되며 수문 분석의 고도화가 이루어지고 있는 오늘날에도, 아직 대수층 깊이 등 지하수관련 매개변수에 대한 연구는 미진한 실정이다. 본 연구는 분포형 모형의 지하수 관련 매개변수 중 지형자료에 해당하는 대수층 깊이의 물리적인 분포형태를 예측하고, 지하수 모의결과를 검토하여 해당 기법의 적용성을 확인하였다. 본 연구에서는 북측의 미계측 유역을 포함한 소양강 유역을 연구대상 지역으로 설정하였고, 정밀한 분포형 모형인 GSSHA(Gridded Surface Hydrologic Analysis)를 활용하였다. 대수층 깊이 추정 방법은 크게 세가지 시나리오로 구분하여 모의를 진행하였다. 유역의 지하수 데이터를 통해 도출된 대수층깊이 등분포(시나리오1), 지표 고도와 대수층 깊이의 선형 반비례 관계를 가정한 선형 회귀식(시나리오2), 동일한 가정을 두고 Log차원에서 회귀식을 적용한 경우(시나리오 3). 위 3가지 시나리오를 통해 산정된 유출량과, 지하수 수위 등을 소양강댐 유입량 자료 및 유역 내 6개 지하수 관측소를 대상으로 결과를 비교하여 적용성을 확인하였다. 시나리오별 유출량 모의 오차평가 결과, 관측 첨두 유량을 가장 잘 반영하고 있는 기법은 일반적으로 선행 연구에서 많이 활용하고 있는 등분포형 기법으로 분석되었으며, 과소·과대 모의된 정도를 나타내는 지표와 모형의 효율성을 나타내는 지표는 선형 회귀분석 기법이 가장 우수한 결과로 분석되었다. 따라서, 대수층 깊이를 등분포하여 모의하던 기존 방식에 비해 지면고도-대수층깊이 간의 반비례 관계를 적용하는 방식이 지하수 모의에 있어서 보다 합리적일 것으로 판단된다. 향후 임의의 인자와 대수층 깊이간의 정밀한 회귀관계를 도출한다면 더욱 합리적이고 신뢰성 높은 결과를 얻을 수 있을것으로 기대된다. 또한 유역 단위의 지하수 모의가 정밀하게 이루어진다면 최근 많은 관심이 집중되는 하천 유지유량과 건기 유출 등의 연구 분야에도 많은 기여를 할 수 있을 것으로 기대된다.

• 한국데이터정보과학회:학술대회논문집
• /
• 2004.04a
• /
• pp.79-82
• /
• 2004

본 발표에서는 1998년 서울대학교 통계학과에서 개발한 KESS(Korean Educational Statistical System; http://stats.snu.ac.kr/time)의 추가로 개발된 내용을 소개하기로 한다 (조신섭 외, 1999). 추가로 개발된 모듈(module)들은 통계교육에서 필요로 하는 분석법들 중에서 회귀분석, 시계열분석과 연관된 내용들이다. 기존의 여러 가지 통계패키지와 비교해 보아 효율적인 통계교육을 위한 필수적인 옵션 및 분석 결과를 제공하도록 하였다.

• Proceedings of the Korean Statistical Society Conference
• /
• 2005.11a
• /
• pp.5-11
• /
• 2005

본 연구에서는 최근에 널리 사용되고 있는 R 프로그램을 이용하여 실험계획법 중 요인배치법과 반응표면분석을 구현하였다. 특히 반응표면분석에서 직교계획, 회전계획, 기울기 회전계획을 만족하는 실험계획을 제공함으로써 상업용 프로그램의 미진한 부분을 개선하여 실험선택의 폭을 넓게 하였다.

• Journal of the Korean Geographical Society
• /
• v.49 no.5
• /
• pp.690-704
• /
• 2014

Single family houses are the traditional & typical type of house in human history. But there had been little attention to single family houses in Korea so that there was little studies on single family houses. This study aimed to analyse price determinants of single family houses in Seoul, using Quantile Regression Analysis(QRA). Because single family houses has large levels of price, quantile regression analysis is more proper than Ordinary Least Square(OLS). The Results of analysis showed that, land coverage ratio, zoning, passed years, basement floor, hight of land, shape of land were important factors to single family houses price. The scale of effect of land coverage ratio to single family houses price was different to price levels of single family houses. And basement floor affected more negative effects to middle price, location and zoning had positive effects to high price single family houses. The degree of influence of determinants of single family houses price was deferent by region, KangBuk and KangNam. In KangNam, land coverage ratio and accessibilities were more important in low price single family houses, green zone and more far way is affected positive effects on single family houses price. In Kangbuk, land coverage ratio affects similar effects on single family houses price.

• The Journal of the Korea Contents Association
• /
• v.12 no.12
• /
• pp.683-693
• /
• 2012

In this study we investigated the factors determining people's decision on whether to subscribe to private health insurance, on how many private health insurances they subscribe to and the average amount of monthly payment from subscribers of private health insurances. For analysis, logistic regression analysis and multiple linear regression analysis were conducted on the sample of 8,167 people using 2008 Korean Longitudinal Study of Ageing(KLoSA) data. From the analysis, whether to enroll in private health insurance is found to be greatly influenced by population and socioeconomic factors as well as regular exercise, smoking, cognitive function scores, subjective health status, hospitalization, the number of outpatient services, free primary health screenings benefits. We also found that number of private health insurances purchased is affected by age, household income, subjective health status, drinking, free primary health screenings benefits and that the average amount of monthly payment for private health insurances purchased is influenced by age, marriage status, economic activities status, subjective sense of hierarchy, household income, drinking, hospitalization. This study is expected to contribute to show the healthy role of private health insurance so that the desirable direction in expansion of health security policy in Korea can be explored further.

• Journal of the Economic Geographical Society of Korea
• /
• v.17 no.1
• /
• pp.129-146
• /
• 2014

This paper investigates how industrial diversity affects unemployment and employment instability from the perspective of the regional economy. Through this analysis, we examine how the industry-specific policy to promote some industry strategically in most of areas affects the stability of the regional economy. We measure Herfindahl indexes using the 1993-2010 data of 16 regions in Korea, and use panel regression model for empirical analysis. The main results from this empirical analysis are summarized as follows. First, we confirm that the industrial structure of most regions has been changed to the direction of specialization in 1990s and to the direction of diversification in 2000s through analyzing the changes in the values of Herfindahl indexes during the given period. Second, we find from the estimation results of panel regression model that the higher industrial diversity in most of regions is, the lower the unemployment rate is. However, a statistically significant relationship between industrial diversity and employment instability only partially confirmed. Third, there exist high unemployment rate and employment instability in most metropolitan areas, but it is hard to say that this relationship is highly statistically significant. From the results of the empirical analysis, it is likely that the industry-specific policies such as the regional strategic industry development policies unlike policy goals make the unemployment rate to rise and economic instability to increase. From the viewpoint of employment aspects, the strategies to increase industrial diversity would be desirable rather than those to specialize in the industrial structure.

• International Area Studies Review
• /
• v.21 no.4
• /
• pp.173-194
• /
• 2017

The research study conducted logistics regression analysis through HLM (Hierarchical Linear Modeling) and presented the value of the outcome in order to investigate characteristic factors of demanders which influence the intention to move into public rental housing. (1) The results of the regression analysis of characteristic factors of household and housing were analyzed as having the significant effect on the intention to move in public rental housing, according to residents moving in monthly rent housing, residents' occupations, rental housing, the number of household, the location of surrounding public rental housing, monthly average income, children's educational level, the number of children, the types of housing and one's own house, in order, out of the types of housing tenure. (2) The results of the regression analysis of characteristic factors of the conditions of location were analyzed that out of the conditions of location of the top five areas in public rental rates, what influences significant effects on the intention to move in public rental housing is the location of surrounding rental housing, income, the number of household and children, children's educational level, job state, housing types, ones' own house, rent housing, monthly rent housing, in order. (3) In case of Seoul, Expanding public rental housing is inevitable in order to stabilize ordinary people's housing stability, owing to the high and rental prices of private housing. Nevertheless, an accurate analysis of the intention to move in public rental housing has not been conducted. Eventually, the research was, thus, conducted, based on the fact that the preference on public housing is low. According to the analytic results of the study, it is required for the government institutions and agencies should consider individual and local characteristics and provide an alternative that meets the real situation, in order to help ordinary citizens with low incomes stabilize housing.