摘要：The valuation of real estates is of extreme importance for decision making. Their singular characteristics make valuation through hedonic pricing methods difficult since the theory does not specify the correct regression functional form nor which explanatory variables should be included in the hedonic equation. In this article, we perform real estate appraisal using a class of regression models proposed by Rigby and Stasinopoulos (2005) called generalized additive models for location, scale and shape (GAMLSS). Our empirical analysis shows that these models seem to be more appropriate for estimation of the hedonic prices function than the regression models currently used to that end..
