Application of Artificial Neural Networks to the Valuation of Residential Property
Peter Rossini · 1997
This paper researches the application of Artificial Neural Networks (ANN) to residential property valuation. Neural Networks (NN) are a multivariate analytical tool which promises to become the next major tool used for computerized mass appraisal. Current literature on ANN is reviewed. The technique is then used in three procedures to establish value estimates. These are compared to results from multiple regression and the actual sale prices. The paper helps to clarify the current position with ANN and seeks to examine its practical use.