Technics of geocoding in construction geographically weighted regression models at the mass estimation in the conditions of uncertainty and heterogeneity of the initial data

Бесплатный доступ

This article considers the application of a method of geographically weighted regress (GWR) to mass estimation of cost of objects in the real estate market. Conditions in which application of the given method allows to receive adequate models of pricing smoothly varying on surveyed territory are defined. The example of concrete construction GWR for the inhabited real estate of Saratov is resulted, comparative analysis of GWR models with model of multiple regress is given. Possibility of use GWR for the analysis of dynamics of price zoning of the market of the real estate of territory is illustrated.

Property market, geographically weighted regress, models with variable structure, modelling, spatially heterogeneous data, mass estimation, price zoning, dynamics of the prices

Короткий адрес: https://sciup.org/170152068

IDR: 170152068

Статья научная