Hedonic Model and Software for Land Pricing and Selecting Key Feature of Land Price

Abstract

      Land pricing and determining the key factors of the land price are useful for land use planning and for the banking. At present, land price estimation is relied on traditional methods. These methods largely depend on analysis and intervention by pricing staff, thus, may not be free of subjective mistakes and non-transparency. In this study, we proposed a method for land pricing based on hedonic model. This model was introduced by Lancaster in 1966 and has many applications in economics including land pricing. Usually, for each region, various particular mathematical models are chosen and tested. In this paper, we investigated a linear model and a second order model in combination with penalty technique to select the key factors affecting land price. The optimal parameters were obtained by solving convex optimization problems. A demo software to calculate the optimal parameters and to evaluate the error was developped. Numerical results with data collected from Long Bien district showed that the proposed method is potentially promising.