• The Korean Journal of Applied Statistics
• /
• v.12 no.2
• /
• pp.605-619
• /
• 1999

본 논문은 회귀분석에서 오차항의 1차 자기상관 존재 여부 및 그 값을 검정하는 방법을 베이지안 접근법으로 제안하였다. 이 방법은 모수공간의 다중분할로 인해 얻어진 여러 가설들에 대한 다중결정문제를 다중 베이즈요인에 관한 이론과 일반화 Savage-Dickey 밀도비를 이용한 사후확률 추정법을 합성하여 개발되었다. 이 방법은 기존의 검정법들에서 가능한 검정 뿐 아니라 이들이 해결할 수 없는 자기상관에 대한 다중결정문제에도 사용이 가능한데 그 효용성이 있다. 모의실험을 통하여 제안된 검정법의 유효성을 평가하였다.

• Journal of the Korean Data and Information Science Society
• /
• v.25 no.4
• /
• pp.807-817
• /
• 2014

This paper considers multivariate time series modelling of PM10 data in Korea collected from 2008 to 2011. We consider both temporal and spatial dependencies of PM10 by applying the sparse vector autoregressive (sVAR) modelling proposed by Davis et al. (2013). It utilizes the partial spectral coherence to measure cross correlation between different regions, in turn provides the sparsity in the model while balancing the parsimony of model and the goodness of fit. It is also shown that sVAR performs better than usual vector autoregressive model (VAR) in forecasting.

• Journal of the Korean Data and Information Science Society
• /
• v.27 no.5
• /
• pp.1203-1214
• /
• 2016

Obesity is a risk factor for various diseases as well as itself a disease and associated with socioeconomic factors. The obesity proportion has been increasing in Korea over about 15 years so that investigation of the socioeconomic factors related with obesity is important in terms of preventation of obesity. In particular, the association between obesity and socioeconomic status varies with gender and has spatial dependency. In the paper, we estimate the effects of socioeconomic factors on obesity proportion by gender, considering the spatial correlation. Here, a conditional autoregressive model under the Bayesian framework is used in order to take into account the spatial dependency. For the real applicaiton, we use the obestiy proportion dataset at 25 districts of Seoul in 2010. We compare the proposed spatial model with a non-spatial model in terms of the goodness-of-fit and prediction measures so the spatial model performs well.

• Journal of Digital Convergence
• /
• v.16 no.10
• /
• pp.89-99
• /
• 2018

The purpose of this study is to analyse the determinants of the change in the welfare expenditure of local governments in 2015. This study analyzed the spatial correlation of welfare expenditure among neighboring local governments and determined the factors affecting the welfare expenditures. According to the results of the study, spatial correlation of welfare expenditure among local governments appears. Determinants, such as socio-economic factors, administrative factors, public financial factors are affecting the amount of the welfare expenditures, but local political factors, and local tax, last year's budgets are not correlated with the amount of local welfare expenditures. In this study, it is significant to found out that the spatial correlation of welfare expenditure among the local governments and to examine the determinants. If possible, it is necessary to analyze the time-series analysis using the multi-year welfare expenditure data, expecially self-welfare expenditures.

• Survey Research
• /
• v.10 no.3
• /
• pp.1-19
• /
• 2009

This study investigates the utilities of spatial analysis in the context of survey research using Geographical Information System(GIS) and Geographically Weighted Regression (GWR) which take account of spatial heterogeneity. Many social phenomena involve spatial dimension, and with the development of GIS, GPS receiver, and online location-based services, spatial information can be collected and utilized more easily, and thus application of spatial analysis in the survey research is getting easier. The traditional OLS regression models which assume independence of observations and homoscedasticity of errors cannot handle spatial dependence problem. GWR is a spatial analysis technique which utilizes spatial information as well as attribute information, and estimated using geographically weighted function under the assumption that spatially close cases are more related than distant cases. Residential survey data from a Primary Autonomous District are used to estimate a model of public service satisfaction. The findings show that GWR handles the problem of spatial auto-correlation and increases goodness-of-fit of model. Visualization of spatial variance of effects of the independent variables using GIS allows us to investigate effects and relationships of those variables more closely and extensively. Furthermore, GIS and GWR analyses provide us a more effective way of identifying locations where the effect of variable is exceptionally low or high, and thus finding policy implications for social development.

• International Area Studies Review
• /
• v.16 no.1
• /
• pp.275-301
• /
• 2012

This study analyzes the regional disparity in India between 24 states over the period 1980 to 2009. The traditional regressive and spatial autoregressive models are used that includes measures of spatial effects. The results provide no evidence that convergence is valid in India. However, the results indicate that spatial interaction is an important element of state growth in India. The result of spatial analysis excluded two outliner states reveals more strong relationship between the weighted spatial income level and the state growth rates. Moreover, the results find that the coefficients of spatial lag of initial per capital and error terms are significantly negative. The coefficient of variation measures that the distribution of state income level has diverged over time. Therefore, this study concludes that the growth of regional state income does not have a tendency to converge rater than diverge. The results is rational because as the Indian economy is growing rapidly, some states grow faster than the others while initial poor states become the poorest ones, which increases regional disparity in India.

• Journal of the Korean Association of Geographic Information Studies
• /
• v.15 no.1
• /
• pp.43-51
• /
• 2012

The OLS(ordinary least squares) method is widely used in hedonic housing models. One of the assumptions of the OLS is an independent and uniform distribution of the disturbance term. This assumption can be violated when the spatial autocorrelation exists, which in turn leads to undesirable estimate results. An alterative to this, spatial econometric models have been introduced in housing price studies. This paper describes the comparisons between OLS and spatial econometric models using housing transaction prices of Busan, Korea. Owing to the approaches reflecting spatial autocorrelation, the spatial econometric models showed some superiority to the traditional OLS in terms of log likelihood and sigma square(${\sigma}^2$). Among the spatial models, the SAR(Spatial Autoregressive Models) seemed more appropriate than the SAC(General Spatial Models) and the SEM(Spatial Errors Models) for Busan housing markets. We can make sure the spatial effects on housing prices, and the reconstruction plans have strong impacts on the transaction prices. Selecting a suitable spatial model will play an important role in the housing policy of the government.

• Journal of the Korean Association of Geographic Information Studies
• /
• v.22 no.1
• /
• pp.1-18
• /
• 2019

This research measured and visualized the spatial dependency and the spatial heterogeneity of the small business in Cheonan-si, Asan-si with $100m{\times}100m$ grids based on global and local spatial autocorrelation. First, we confirmed positive spatial autocorrelation of small business in the research area using Moran's I Index, which is ESDA(Exploratory Spatial Data Analysis). And then, through Getis-Ord $GI{\ast}$, one kind of LISA(Local Indicators of Spatial Association), local patterns of spatial autocorrelation were visualized. These verified that Spatial Regression Model is valid for the location factor analysis on small business commercial buildings. Next, GWR(Geographically Weighted Regression) was used to analyze the spatial relations between the distribution of small business, hourly mobile traffic-based floating population, land use attributes index, residence, commercial building, road networks, and the node of traffic networks. Final six variables were applied and the accessibility to bus stops, afternoon time floating population, and evening time floating population were excluded due to multicollinearity. By this, we demonstrated that GWR is statistically improved compared to OLS. We visualized the spatial influence of the individual variables using the regression coefficients and local coefficients of determinant of the six variables. This research applied the measured population information in a practical way. Reflecting the dynamic information of the urban people using the commercial area. It is different from other studies that performed commercial analysis. Finally, this research has a differentiated advantage over the existing commercial area analysis in that it employed hourly changing commercial service population data and it applied spatial statistical models to micro spatial units. This research proposed new framework for the commercial analysis area analysis.

• Journal of the Korean Data and Information Science Society
• /
• v.28 no.1
• /
• pp.11-20
• /
• 2017

In this paper, we introduced linear spatial time series (space-time autoregressive and moving average model) and nonlinear spatial time series (space-time bilinear model). Also we estimated the parameters by Kalman Filter method and made comparative studies of power of forecast in the final model. We proposed several weight matrices such as equal proportion allocation, reciprocal proportion between distances, and proportion of population sizes. For applications, we collected Mumps data at Korea Center for Disease Control and Prevention from January 2001 until August 2008. We compared three approaches of weight matrices using the Mumps data. Finally, we also decided the most effective model based on sum of square forecast error.

• Journal of the Korean Society of Surveying, Geodesy, Photogrammetry and Cartography
• /
• v.37 no.3
• /
• pp.129-141
• /
• 2019

In recent years, fast growing cities in Korea are showing signs of being vulnerable to more disasters as their population and facilities increase and intensify. In particular, fire is one of the most common disasters in Korea's cities, along with traffic accidents. Therefore, in this study, we analyze what type of factors affect the fire that threatens urban people. Fire data were acquired for 10 years, from 2007 to 2017, in Jinju, Korea. Spatial distribution pattern of fire occurrence in Jinju was assessed through the spatial autocorrelation analysis. First, spatial autocorrelation analysis was carried out to grasp the spatial distribution pattern of fire occurrence in Jinju city. In addition, correlation and multiple regression analysis were used to confirm spatial dependency and abnormality among factors. Based on this, OLS (Ordinary Least Square) regression analysis was performed using space weighting considering fire location and spatial location of each facility. As a result, First, LISA (Local Indicator of Spatial Association) analysis of the occurrence of fire in Jinju shows that the most central commercial area are fire department, industrial area, and residential area. Second, the OLS regression model was analyzed by applying spatial weighting, focusing on the most derived factors of multiple regression analysis, by integrating population and social variables and physical variables. As a result, the second kind of neighborhood living facility showed the highest correlation with the fire occurrence, followed by the following in the order of single house, sales facility, first type of neighborhood living facility, and number of households. The results of this study are expected to be useful for analyzing the fire occurrence factors of each facility in urban areas and establishing fire safety measures.