Developing a multi-network urbanization model: A case study of urban growth in Denver, Colorado

摘要: Urbanization is an important issue concerning diverse scientific and policy communities. Computational models quantifying locations quantities of urban growth offer numerous environmental socioeconomic benefits. Traditional are based on a single-algorithm fitting procedure thus restricted their ability to capture spatial heterogeneity. Accordingly, GIS-based modeling framework titled multi-network urbanization (MuNU) model developed that integrates multiple neural networks. The MuNU enables filtering approach where input data patterns automatically reallocated into appropriate networks with targeted accuracies. We hypothesize observations classified by individual share greater homogeneity, accuracy will increase the integration algorithms. Land use land cover sets two time snapshots (1977 1997) covering Denver Metropolitan Area used for training validation. Compared single-step algorithm-either stepwise logistic regression or single network-several improvements evident in visual output model. Statistical validations further quantify superiority support our hypothesis effective incorporation