Action Filename Size Access Description License
Show more files...


The real-estate market is "where" a multiplicity of economic, cultural, social and demographic factors are synthesised with respect to choices regarding the qualitative and locational aspects of a property. The spatial analysis of the real-estate market and, in particular, of the factors which contribute to determining prices, is a very useful instrument in outlining the geography of the economic development of vast areas. The aim of the paper is the construction of a simulation model, on a spatial level, of real-estate values with reference to the housing market in the urban area of the city of Treviso (I). The model was built using a neural network which gives the possibility of analysing the marginal contribution of single real-estate characteristics independently of the a priori choice of the interpolation function; at the same time it works well even in the presence of statistical correlation among the explicative variables, a serious drawback in multiple regression models. The work is divided into several parts. First, a synthetic picture of the real-estate market of the area studied has been drawn up with reference to the main conditioning factors. Then the problem of the selection of a neural network model for the appraisal of property values is presented. Finally, there is the description of the procedure for the spatialization of obtained results from the neural model for the definition of a values map. The results shows the notable interpretative and predictive capacity of the neural model and it seems very useful in appraisals. Furthermore, the mapping of value fluctuations enables first-hand verification of the "goodness" of the assessed model and its capacity to portray the real situation. The general approach presented seems, therefore, useful both as an instrument of support for urban and territorial planning, as well as a permanent monitoring system of the real-estate market with the aim of creating an informative system of support for the analysis of real-estate investment.


Downloads Statistics

Download Full History