Estimation of Housing Price Variations Using Spatio-Temporal Data

Abstract

This paper proposes a hedonic regression model to estimate housing prices and the spatial variability of prices over multiple years. Using the model, maps are obtained that represent areas of the city where there have been positive or negative changes in housing prices. The regression-cokriging (RCK) method is used to predict housing prices. The results are compared to the cokriging with external drift (CKED) model, also known as universal cokriging (UCK). To apply the model, heterotopic data of homes for sale at different moments in time are used. The procedure is applied to predict the spatial variability of housing prices in multi-years and to obtain isovalue maps of these variations for the city of Granada, Spain. The research is useful for the fields of urban studies, economics, real estate, real estate valuations, urban planning, and for scholars.

A comparative study of land price estimation and mapping using regression kriging and machine learning algorithms across Fukushima prefecture, Japan

“…The models' prediction errors are relatively equal, which range approximately between 15.1% and 15.9% for validation and cross-validation. The exponential model gives slightly better outcomes in both tests, which coincide with the result found by Tsutsumi et al (2011) as well as Chica-Olmo et al (2019). Figure 7 shows the land price maps for the year 2015 compiled based on officially published observational data.…”

“…To assess the estimation accuracy, we performed validation and cross-validation tests based on randomly selected training and testing samples representing 70% and 30% of the samples, respectively. Both tests showed that the exponential model (krig.EXP) gives slightly better results than the Gaussian model (krig.GAU) and the spherical model (krig.SPH), which concur with the conclusions reached by Tsutsumi et al (2011) and Chica-Olmo et al (2019). The second approach for predicting land prices was to employ nine popular ML algorithms split into three categories: (1) linear (GLM, GAMS, and SVMLinear), (2) nonlinear (MARS, kNN, and SVMRadial), and (3) regression tree (Cubist, GBM, and RF).…”

“…Kriging and cokriging models have been implemented for different topics in geosciences (e.g., remote sensing, climatology, and agriculture) and have shown robust performance in terms of estimation errors. Due to their flexibility, Palma et al (2019) affirmed the capability of geostatistical methods to tackle socioeconomic phenomena as well, which can be seen in the number of studies conducted to model land or house rental prices (e.g., Chica-Olmo, 2007;Chica-Olmo et al, 2019;Kuntz and Helbich, 2014;Luo and Wei, 2004;Tsutsumi et al, 2011). Most recently, new studies introduced machine learning (ML) algorithms as possible alternatives to hedonic models and geostatistical methods to describe spatially and temporally the distribution of land prices.…”

“…Adegoke found a significant relationship between land price and mainly the population density and level of development. Other studies have linked proximity to urban facilities, such as schools, parks, and sports-related venues with mainly residential prices in Singapore (Murakami, 2018), China (Liu et al, 2007), Kuwait (Mostafa, 2018), Spain (Chica-Olmo et al, 2019), and the United States (Clapp et al, 2008;Espey and Owusu-Edusei, 2001; Kiel and Zabel, 2008). In conclusion, the selection process of suitable factors depends mainly on various elements, such as the setting of the designated target area, type of land price (e.g., residential and commercial), and the availability of aspatial or spatial data.…”

Population Aging, Household Savings and Asset Prices: A Study Based on Urban Commercial Housing Prices

“…In the literature, many studies have investigated the effects of the aging population [1][2][3][4]. In addition, many studies have investigated housing prices [5][6][7], and these studies have laid the foundation for our research.…”