• Communications for Statistical Applications and Methods
• /
• 제13권1호
• /
• pp.177-190
• /
• 2006

The aim of this study is to propose a Bayesian model for fitting mortality rate of colon cancer. For the analysis of mortality rate of a disease, factors such as age classes of population and spatial characteristics of the location are very important. The model proposed in this study allows the age class to be a random effect in addition to its conventional role as the covariate of a linear regression, while the spatial factor being a random effect. The model is fitted using Metropolis-Hastings algorithm. Posterior expected predictive deviances, standardized residuals, and residual plots are used for comparison of models. It is found that the proposed model has smaller residuals and better predictive accuracy. Lastly, we described patterns in disease maps for colon cancer.

• 한국항해항만학회지
• /
• 제29권5호
• /
• pp.431-437
• /
• 2005

화물발생 예측을 위해 주로 이용된 방법은 교통존을 중심으로 한 공간단위의 접근기법이다. 또한 연구사례에 따라 물동량 전수화 방법에 차이가 있고 모형에 포함된 변수가 다른 경우가 많아 일관된 화물발생 예측에 어려움이 있다. 본 연구에서 시도한 물동량 단위 예측기법은 개별 사업체를 대상으로 유출${\cdot}$입하는 화물 물동량과 종업원수, 매출액, 연면적, 부지면적 등 사업체의 특성을 이용하여 사업체 단위로 화물발생을 예측하는 것으로 공간단위 접근기법보다 간편하고 다른 지역에 적용이 가능한 장점이 있다. 연구의 대상은 수도권에 소재한 사업체이며 매출액을 설명변수로 선정하여 회귀식을 추정하고 모형의 적합성을 검증하였다. 회귀모형의 형태는 지수형으로 매출액이 증가함에 따라 화물발생의 규모도 커지는 특징을 갖는다. 본 연구는 교통 존 단위가 아닌 사업체 단위로 화물발생을 예측하는 보다 미시적인 연구방법을 개발함에 의의가 있다.

• 한국지구과학회지
• /
• 제31권5호
• /
• pp.454-464
• /
• 2010

A probability prediction model for tropical cyclone (TC) genesis in the Northwestern Pacific area was developed using the logistic regression method. Total five predictors were used in this model: the lower-level relative vorticity, vertical wind shear, mid-level relative humidity, upper-level equivalent potential temperature, and sea surface temperature (SST). The values for four predictors except for SST were obtained from difference of spatial-averaged value between May and January, and the time average of Ni$\tilde{n}$o-3.4 index from February to April was used to see the SST effect. As a result of prediction for the TC genesis frequency from June to December during 1951 to 2007, the model was capable of predicting that 21 (22) years had higher (lower) frequency than the normal year. The analysis of real data indicated that the number of year with the higher (lower) frequency of TC genesis was 28 (29). The overall predictability was about 75%, and the model reliability was also verified statistically through the cross validation analysis method.

• 한국지리정보학회지
• /
• 제22권1호
• /
• pp.1-18
• /
• 2019

본 연구에서는 천안시, 아산시의 $100m{\times}100m$격자 내에 집계된 소상공인 분포가 가지는 공간적 의존성(Spatial Dependency)과 공간적 이질성(Spatial Heterogeneity)을 전역적(Global), 국지적(Local) 공간 자기상관(Spatial Autocorrelation)을 통해 측정 및 가시화하였다. 먼저 탐색적 공간데이터 분석방법(ESDA: Explotory Spatial Data Analysis)인 Moran's I Index를 통해 연구지역에서 소상공인 분포의 정적(Positive)공간자기상관이 발생하는 것을 확인하였으며, 국지적 공간자기상관 지표(LISA : Local Indicators of Spatial Association) 중 하나인 Getis-Ord $GI{\ast}$를 통해 공간자기상관의 국지적 패턴을 가시화하였다. 이를 통해 소상공인 상가점포의 입지요인 분석 시 적용할 변수와의 관계에 대해 공간회귀모형의 적용이 타당함을 증명하였으며, 소상공인의 분포와 모바일 트래픽 기반의 시간대별 유동인구, 토지이용 혼합성 지수 그리고 주거지, 상점, 도로망, 교통결절점과의 공간영향관계를 지리가중 회귀분석(GWR : Geographically Weighted Regression)을 통해 분석하였다. 최종적으로 다중공선성(Multicollinearity)이 발생했던 버스정류장 접근성, 오후시간대 유동인구, 저녁시간대 유동인구를 제외한 6개의 변수를 적용하였고 GWR 모형이 OLS모형보다 주요통계량에서 모형 설명력이 개선됨을 도출하였다. 분석에 최종적으로 적용된 6가지 변수의 회귀계수와 국지적 결정계수(Local $R^2$)에 대해 연구지역 내에서 공간적으로 변화하는 변수별 영향력을 가시화하였다. 본 연구는 실질적으로 측정된 방식의 유동인구 정보를 적용함으로써 상권을 이용하는 도시민의 동적 정보를 반영한 것이 상권분석을 수행한 다른 연구들과 차별적인 성격을 가진다. 마지막으로 이러한 동적정보와 변수들의 공간적 상호작용을 구조화하기 위해 미시적 공간단위에서 공간통계학(Spatial Statistical)적 모형 적용을 통해 상권분석의 새로운 프레임을 제시하였다는 점에서 연구적 의의를 가진다.

• 한국지형학회지
• /
• 제23권3호
• /
• pp.67-78
• /
• 2016

Soil carbon(C) is an essential property for characterizing soil quality. Understanding spatial patterns of soil C is particularly limited for mountain areas. This study aims to predict the spatial pattern of soil C using terrain analysis in a steep mountainous area. Specifically, model performances and prediction uncertainties were investigated based on the number of resampling repetitions. Further, important predictors for soil C were also identified. Finally, the spatial distribution of uncertainty was analyzed. A total of 91 soil samples were collected via conditioned latin hypercube sampling and a digital soil C map was developed using support vector regression which is one of the powerful machine learning methods. Results showed that there were no distinct differences of model performances depending on the number of repetitions except for 10-fold cross validation. For soil C, elevation and surface curvature were selected as important predictors by recursive feature elimination. Soil C showed higher values in higher elevation and concave slopes. The spatial pattern of soil C might possibly reflect lateral movement of water and materials along the surface configuration of the study area. The higher values of uncertainty in higher elevation and concave slopes might be related to geomorphological characteristics of the research area and the sampling design. This study is believed to provide a better understanding of the relationship between geomorphology and soil C in the mountainous ecosystem.

• 지식경영연구
• /
• 제17권3호
• /
• pp.45-64
• /
• 2016

As the digital retail environement becomes prevalent, consumers are given greater opportunities to make purchases across physical and digital boundaries. Prior research emphasizes that the attractiveness of the digital or online channel is relatively determined by spatial specifics of physical locations. The overall market trend combined with prior research suggests that understanding spatial specifics becomes a key to managing both offline and online sales performance together. In this study, we focus on geographic variation in sales performance through offline and online channels and aim to investigate the channel-level sales difference between central and subsidiary areas. To this end, we obtain sales data of skincare and makeup products from a leading cosmetic company. Next, we examine spatial autocorrelations in data and then employ the spatial error models to study the effects of spatial specifics. The empirical findings are as follows. First, there are significant differences in category-specific and channel-level sales between central and subsidiary areas. Second, Moran's I statistics demonstrate the spatial autocorrelations of each variable. Third, spatial error models outperform simple regression models with lower AIC values. Finally, spatial specifics play a greater role in understanding online sales in subsidiary areas whereas they exert greater influence on offline sales in central areas. We believe our study advances the related theory and knowledge of multi-channel retailing and also contributes practically to location-dependent multi-channel strategies and sales data analytics.

• 대한공간정보학회지
• /
• 제21권4호
• /
• pp.73-81
• /
• 2013

식생의 분포와 지형 태양광도의 상관성을 규명하는 것은 공간적 이질성을 내포하는 공간데이터의 분석이지만 기존의 많은 선형모델들은 이들 데이터가 갖는 공간적 특성을 고려하지 못하고 있다. 이러한 문제점을 극복하기 위해 금강산을 대상으로 식생분포를 정량적으로 나타내는 NDVI(Normalized Difference Vegetation Index)와 일사량, 일조시간, 고도, 경사에 대하여 지리가중회귀분석(GWR : Geographically Weighted Regression)을 실시하였다. GWR 은 전역적 모형인 OLS(Ordinary Least Squares)에 비해 모형의 설명력과 적합성이 확연히 높아졌으며, 잔차의 공간적 자기상관성 또한 해소된 것으로 나타났다. OLS 분석결과는 NDVI에 미치는 지형 태양광도의 영향력을 연구지역에서 단일하게 추정하였으나, GWR은 각 인자가 NDVI에 미치는 영향력을 국지적으로 보다 세밀하게 추정하여 공간단위에 따른 각 인자의 영향력을 보다 확연히 나타내었다. 국지적 차원에서 추정된 NDVI와 지형 태양광도의 상관성은 식생분포를 조사하는 과정에서 보다 객관적이고 세밀한 분석을 위한 중요한 참고자료로 사용될 수 있을 것이다.

• 보건행정학회지
• /
• 제23권4호
• /
• pp.387-396
• /
• 2013

Background: The purpose of this study was to analyze the association between areas of Korea Train Express (KTX) region and external medical service use in Korean society using spatial statistical model. Methods: The data which was used in this study was extracted from 2011 regional health care utilization statistics and health insurance key statistics from National Health Insurance Corporation. A total spatial units of 229 districts (si-gun-gu) were included in this study and spatial area was all parts of the country excepted Jeju, Ulleungdo island. We conducted Kruskal-Wallis test, correlation, Moran's I and hot-spot analysis. And after, ordinary linear regression, spatial lag, spatial error analysis was performed in order to find factors which were associated with external medical service use. The data was processed by SAS ver. 9.1 and Geoda095i (windows). Results: Moran's I of health insurance patients' external medical service use was 0.644. Also, population density, Seoul region, doctor factors positively associated with health insurance patients' external medical service. In contrast, average age, health care organization per 100 thousand were negatively associated with health insurance patients' external medical service use. Conclusion: The finding of this study suggested that health insurance patient's external medical service use correlated for seoul region in korea. The study results imply the need for more attention medical needs in the region (si-gun-gu unit) for health insurance patients of seoul region. It is important to adapt strategy to activation of primary health care as well as enhancing public health institution for prevent leakage of patients to other areas.

• 국제초고층학회논문집
• /
• 제5권2호
• /
• pp.95-103
• /
• 2016

One of the key problems in the design of high-rise commercial complex is how to guide reasonable pedestrian distribution in commercial space. In this study, pedestrian distribution in three high-rise commercial complexes in Shanghai and Hong Kong was studied using spatial configuration analysis software Space Syntax and quantification of physical elements in commercial spaces, such as functional attractiveness, entrances, escalators, level variations and passage width. Additionally, in an attempt to integrate functions with spatial integration and spatial depth, two combination variables, the spatial coefficient of function (IF) and spatial depth coefficient of function (F/D), were proposed. The results of the correlation analysis and multiple regression analyses reflected the following: (1) Regarding the influence on pedestrian distribution, there was a synergistic and complementary relationship between function and space; (2) The comprehensive flow distribution analytic model could successfully interpret flow distribution in high-rise commercial complexes and its R Square ranged up to about 70% in the three cases; (3) The spatial coefficient of function (IF) and spatial depth coefficient (F/D) could effectively integrate functions and spatial configuration, which could help close the gap between over-emphasis on function in commercial research and the lack of consideration of function in space-syntax analysis.

• 한국농공학회논문집
• /
• 제59권1호
• /
• pp.11-20
• /
• 2017

This study is to estimate the spatial soil moisture using multiple linear regression model (MLRM) and 15 minutes interval Land Surface Temperature (LST) data of Communication, Ocean and Meteorological Satellite (COMS). For the modeling, the input data of COMS LST, Terra MODIS Normalized Difference Vegetation Index (NDVI), daily rainfall and sunshine hour were considered and prepared. Using the observed soil moisture data at 9 stations of Automated Agriculture Observing System (AAOS) from January 2013 to May 2015, the MLRMs were developed by twelve scenarios of input components combination. The model results showed that the correlation between observed and modelled soil moisture increased when using antecedent rainfalls before the soil moisture simulation day. In addition, the correlation increased more when the model coefficients were evaluated by seasonal base. This was from the reverse correlation between MODIS NDVI and soil moisture in spring and autumn season.