The contribution is focused on comparing different automated valuation models dealing in different ways with location variable. Using a sample of 290 observations in Minsk, several different AVM models will be compared. A linear and log linear model of AVM with constant location variable will be applied. Therefore, after the determination of spatial correlation using the Moran I test the application of mixed regressive model integrating the geographic variable with the specific technical characteristics of the property. The results confirm an increasing quality of the model. Further works can be required to include also temporal variable in this class of models (Borst 2015).
Dealing with Spatial Modelling in Minsk / D’Amato, Maurizio; Siniak, Nikolaj; Amoruso, Paola (STUDIES IN SYSTEMS, DECISION AND CONTROL). - In: Advances in Automated Valuation Modeling: AVM After the Non-Agency Mortgage Crisis / Maurizio d'Amato, Tom Kauko ; [a cura di] Maurizio d'Amato and Tom Kauko. - Cham, CH : Springer, 2017. - ISBN 978-3-319-49744-0. - pp. 201-208 [10.1007/978-3-319-49746-4_12]
Dealing with Spatial Modelling in Minsk
Maurizio d’Amato;
2017-01-01
Abstract
The contribution is focused on comparing different automated valuation models dealing in different ways with location variable. Using a sample of 290 observations in Minsk, several different AVM models will be compared. A linear and log linear model of AVM with constant location variable will be applied. Therefore, after the determination of spatial correlation using the Moran I test the application of mixed regressive model integrating the geographic variable with the specific technical characteristics of the property. The results confirm an increasing quality of the model. Further works can be required to include also temporal variable in this class of models (Borst 2015).I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
