GIS Spatial Regression Analysis of Land Price Variation -An Example of Eastern Tainan City

• Main Authors: Kung-TingChen, 陳冠廷
• Other Authors: Ming-Chee Wu
• Format: Others
• Language: zh-TW
• Published: 2013
• Online Access: http://ndltd.ncl.edu.tw/handle/r7w6yt

碩士 === 國立成功大學 === 地球科學系專班 === 101 === A lot of quantitative research in the natural and social sciences today is based on a single subject. We typically use the consequences of a single subject to expand it into an equation that is built by an explanatory variable. The one we usually see will be called “Ordinary Least Square” (OLS). The social sciences basically focus on how to describe the connection between single consequences and group explanatory variables. The prices of immovable property often use Hedonic Price Analysis and equipment to find out what will make the price of the land change. This research try to build a connection between dependent and independent variables, no matter whether there are any consequences. When we start a project, there is already a hint that location and space will not affect it much, but the fact is that every single variable has an effect to some degree. This effect will decrease as the distance increase. This research is based on three different types of spatial models: spatial lag model, spatial error model and ordinary least squares. We use the real estate prices in the eastern district of Tainan City, Taiwan R.O.C. between the first season of 2009 to the fourth season of 2010 This study will analyze how the prices in the eastern district of Tainan City changed using Spatial Econometrics, as well as the global spatial dependency test and local indicators of spatial association (LISA). Then, the advantages and weaknesses between the spatial lag model and spatial error model will be analyzed, prove that the experimental area most suitable mode is spatial error model, the price of land clusters tend to structure effects of the location.