Abstract
Analyzing Southern California’s Housing Market Using GIS and Neural Networks
Track: Business GIS
Author(s): Carsten Lange
The paper identifies and quantifies the contributing factors that have an impact on residential housing prices in the research area using a GIS and neural networks approach. A neural network is a computational system to model complex relationships or to find patterns in data and is especially suited for non-linear relationships.
The underlying hedonic model is based on a combination of Geographical Information Systems and Neural Networks to identify and quantify significant variables that contribute to a change in residential housing prices.
The model was developed in two steps. First, a real estate database was developed containing transaction prices for residential housing in recent years as well as spatial and non-spatial attributes that can potentially influence housing prices. In a second step the database was used as an input for a neural network to identify significant variables and quantify their contribution to housing price changes in the selected research area.
Carsten Lange
California State University
3801 W Templer Ave
Pomona , California 91768
United States
Phone: (909) 869 3843
E-mail: clange@csupomona.edu