The powerful pattern-recognition properties of Artificial Neural Networks (ANNs) are outlined and shown to discriminate the effect of aircraft noise from other value determinants in a study of residential property values adjacent to Manchester International Airport. The varying effects of noise on different property types and neighbourhoods are identified and examined. The results are contrasted with earlier hedonic regression-based studies, particularly that of Pennington Topham and Ward (1990) which was the source of the data analysed. (Author/publisher).
Abstract