Neural networks for the mass appraisal of real estate: Fid-Bau Portal

Neural networks for the mass appraisal of real estate

Kathmann, Ruud M.

AbstractIn this paper, a method for mass appraisal is presented which is based entirely on the estimation of value differences between properties. The value difference between two properties is calculated by multiplication of the differences in the characteristics considered, with marginal adjustment factors for those characteristics. The marginal adjustments used in this method are calculated using artificial neural networks. Artificial neural networks are special computer programs which establish a network of very simple operating units. Each unit transforms its input according to the same function, resulting in an output. The input can come from outside the network (input unit) or from other units within the network. The amount to which the output of one unit contributes to the input of another unit is controlled by the weights. The definition of these weights is done by “training ” the network. The artificial neural network for appraisal can be trained with “patterns” of characteristics of properties and the market value of those properties. The training of the network can be considered as the market analysis with the available sales data. Market analysis is an essential part of each mass-appraisal process. The market analysis using neural networks provides substantially more data to the appraisal process than other methods of market analysis.