• Journal of Biosystems Engineering
• /
• 제39권3호
• /
• pp.151-157
• /
• 2014

Purpose: In order to develop strategies to prevent farm-work accidents relating to agricultural machinery, influential factors were examined in this paper. The effects of these factors were quantified using logistic regression. Methods: Based on the results of a survey on farm-work accidents conducted by the National Academy of Agricultural Science, 21 tentative independent variables were selected. To apply these variables to regression, the presence of multicollinearity was examined by comparing correlation coefficients, checking the statistical significance of the coefficients in a simple linear regression model, and calculating the variance inflation factor. A logistic regression model and determination method of its goodness of fit was defined. Results: Among 21 independent variables, 13 variables were not collinear each other. The results of a logistic regression analysis using these variables showed that the model was significant and acceptable, with deviance of 714.053. Parameter estimation results showed that four variables (age, power tiller ownership, cognizance of the government's safety policy, and consciousness of safety) were significant. The logistic regression model predicted that the former two increased accident odds by 1.027 and 8.506 times, respectively, while the latter two decreased the odds by 0.243 and 0.545 times, respectively. Conclusions: Prevention strategies against factors causing an accident, such as the age of farmers and the use of a power tiller, are necessary. In addition, more efficient trainings to elevate the farmer's consciousness about safety must be provided.

• 한국임상보건과학회지
• /
• 제1권1호
• /
• pp.9-17
• /
• 2013

Purpose. This study is to find the degree of variations and trends of hospital services utilization for hypertensive disease, and have done the comparative analysis of the factors affecting occurring some variations. For this, this study uses the data for patients-survey and health-survey of a regional society by Korea Institute for Heath and Social Affairs in 2008; The regional units are classified into 160 of medium size medical service areas. Methods. I understand the level of variation by using index of Extremal Quotient(EQ) and Coefficient Variation(CV), and analyze critical factors influencing some differences in hospital services utilization by using multi-regression model. Results. The main results are followed:The first, in case of rate of hospital services utilization according to standarization of sex and age by small area, I find the variations of EQ 5.3 and CV 0.3; In Ho-nam, especially, the variation of high rank of 10 of age shows higher distribution. The second, the results analyzing the factors influencing on hospital services utilization by multi regression model are that a number of bed hospitals is significant positive relationship and EQ-5D of health behavior is significant negative one. Conclusions. To increase equity of hospital services utilization for hypertensive disease, this study requests the appropriate supply management of bed hospitals by region, efficient allocation of resources, and revitalization of the health promotion program.

• Communications for Statistical Applications and Methods
• /
• 제11권3호
• /
• pp.583-595
• /
• 2004

For the Korean Economically Active Population Survey(EAPS), we consider the composite estimator based on logistic regression model to estimate the unemployment rate for small areas(Si/Gun). Also, small area estimation technique based on hierarchical generalized linear model is proposed to include the random effect which reflect the characteristic of the small areas. The proposed estimation techniques are applied to real domestic data which is from the Korean EAPS of Choongbuk. The MSE of these estimators are estimated by Jackknife method, and the efficiencies of small area estimators are evaluated by the RRMSE. As a result, the composite estimator based on logistic model is much more efficient than others and it turns out that the composite estimator can produce the reliable estimates under the current EAPS system.

• 응용통계연구
• /
• 제31권3호
• /
• pp.329-342
• /
• 2018

노모그램은 질병의 위험 요인과 예측 확률을 쉽게 이해할 수 있도록 시각적으로 표현하는 통계적 도구이다. 본 논문은 만성 폐쇄성 폐질환(chronic obstructive pulmonary disease)의 위험 요인을 이용하여 로지스틱 회귀모형과 순수 베이지안 분류기 모형의 노모그램을 구축하고 이를 비교하였다. 분석 데이터는 국민건강영양조사 6기(2013-2015)를 이용하여 진행하였다. 총 6개의 위험 요인을 이용하였다. 그리고 로지스틱 회귀모형, 순수 베이지안 분류기 모형과 각각의 구축 방법을 이용하여 만성 폐쇄성 폐질환의 노모그램을 제시하였다. 또한, 구축된 두 노모그램을 비교하여 유용성을 살펴보았다. 마지막으로 ROC curve와 Calibration plot을 통하여 각 노모그램을 검증하였다.

• 한국학교ㆍ지역보건교육학회지
• /
• 제23권4호
• /
• pp.81-93
• /
• 2022

Objectives: The purpose of this study was to develop a model for comprehensively evaluating the risk of sarcopenia in Korean adults and to generate the sarcopenia risk scorecard model based on the results. Methods: The participants of the study were 7,118 adults without sarcopenia in the first basic survey, and a longitudinal analysis was conducted using data from the 1st to 8th survey (2006-2020) of the Korean Longitudinal Study of Aging (KLoSA). The data were analyzed using Rao-Scott chi-square test and weighted Cox proportional hazards regression of complex sampling design. The sarcopenia risk scorecard model was developed by Cox proportional hazards regression using points to double the odds (PDO) method. Results: The findings show that the risk factors for sarcopenia in Korean adults were gender, age, marital status, socioeconomic status, body mass index (BMI), regular exercise, diabetes and arthritis diagnosis. In the scorecard results, the case of exposure to the highest risk level was 100 points. The highest score range were given in the order of age over 65, low BMI, and low socioeconomic status. Conclusions: The significance of this study is that the causal relationship between various factors and the occurrence of sarcopenia in Korean adults was identified. Also, the model developed in this study is expected to be useful in detecting participants with risk of sarcopenia in the community early and preventing and managing sarcopenia through appropriate health education.

• 지질공학
• /
• 제14권2호
• /
• pp.211-222
• /
• 2004

이 연구는 자연사면에서 발생하는 토석류(debris flow)산사태의 확률론적 예측을 위해 로지스틱 회귀분석(logistic regression analysis)을 이용하여 변성 암 및 화강암 분포지에 적용할 수 있는 예측모델을 개발한 것이다. 산사태 예측모델을 개발하기 위해 경기 남ㆍ북부지역과 경북 상주지역에서 발생한 산사태 자료를 현장조사와 실내토질시험을 통해 직접 획득ㆍ분석하였다. 산사태 발생에 영향을 미치는 인자는 기초 통계분석은 물론 로지스틱 회귀분석을 실시하여 최종적으로 7개 영향인자를 선정하였다. 이들 7개 인자는 지형요소 2개와 지질 및 토질특성 요소 5개로 구성되어 있고, 각 인자별 가중치를 부여한 점이 큰 특징이다. 개발된 모델은 신뢰성 검증을 수행한 결과 90.74%의 예측율을 확보한 것으로 나타났다. 이 모델을 이용하여 산사태 발생가능성을 확률적ㆍ정량적으로 예측할 수 있게 되었다.

• 대한토목학회논문집
• /
• 제36권3호
• /
• pp.427-437
• /
• 2016

본 연구에서는 지형 및 기상학적 인자만으로 미계측 유역의 저유량부 유황곡선을 추정할 수 있는 지역회귀모형을 개발하고자 하였다. 이를 위해서 16개 유역의 계측 자료로부터 저유량 영역(지속일수 185일에서 365일)에 대한 유황곡선을 작성하고, 이를 토대로 로그형태의 이변수 회귀모형을 구축하였다. 이 회귀모형을 미계측 유역에 적용할 수 있도록 유역면적, 유역경사, 수계밀도, 연평균강수량, 연평균유출량, 유출곡선지수 등의 유역특성인자를 이용하여 모형의 매개변수를 지역화 하였다. 개발한 지역회귀모형으로 평균갈수량, 평균저수량, 평균평수량을 추정하여 관측값과 비교한 결과, 유역면적, 유출곡선지수, 연평균강수량 조합으로 구성된 지역회귀모형이 가장 우수한 것으로 분석되었다.

• Communications for Statistical Applications and Methods
• /
• 제18권1호
• /
• pp.47-55
• /
• 2011

비대체와 회귀대체는 조사변수의 모형과 조사변수와 보조변수의 관계에 의존하며 모형이 성립되지 않는 경우 이들 대체법을 이용한 추정량의 불편성은 보장되지 않는다. 본 연구에서는 모형이 성립되지 않는 경우에도 추정량의 근사적 불편성이 성립되는 로버스트 대체법을 개발한다. 대체법 개발시 보조변수의 모수 정보를 이용하여 추정량의 효율 증대를 가져오게 한다. 모의실험을 실시하여 본 연구에 대한 이론적 결과의 타당성을 보인다.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제8권6호
• /
• pp.2056-2069
• /
• 2014

In order to implement interactive and personalized Web services properly, it is necessary to understand the tangible and intangible responses of the users and to recognize their emotional states. Recently, some studies have attempted to build emotional state estimation models based on facial expressions. Most of these studies have applied multiple regression analysis (MRA), artificial neural network (ANN), and support vector regression (SVR) as the prediction algorithm, but the prediction accuracies have been relatively low. In order to improve the prediction performance of the emotion prediction model, we propose a novel SVR model that is optimized using a genetic algorithm (GA). Our proposed algorithm-GASVR-is designed to optimize the kernel parameters and the feature subsets of SVRs in order to predict the levels of two aspects-valence and arousal-of the emotions of the users. In order to validate the usefulness of GASVR, we collected a real-world data set of facial responses and emotional states via a survey. We applied GASVR and other algorithms including MRA, ANN, and conventional SVR to the data set. Finally, we found that GASVR outperformed all of the comparative algorithms in the prediction of the valence and arousal levels.

• 응용통계연구
• /
• 제24권3호
• /
• pp.445-453
• /
• 2011

Almost all small area estimations are obtained by minimizing the mean squared error. Recently relative error prediction methods have been developed and adapted to small area estimation. Usually the estimators obtained by using relative error prediction is called a shrinkage estimator. Especially when data set consists of large range values, the shrinkage estimator is known as having good statistical properties and an easy interpretation. In this paper we study the shrinkage estimators based on logistic regression type estimators for small area estimation. Some simulation studies are performed and the Economically Active Population Survey data of 2005 is used for comparison.