Atlanta Smart Growth

Geo-spatial machine learning in the land use modeling process 

For this project, considerable amount of vector and raster data is collected from USGS, Census and open data websites, and wrangled together into a model-ready dataset using ArcGIS & RStudio.

A Binary Logistic Regression Model is constructed and trained based on the change in development from 2001-2011. After cross validation, the model predicts the probability of development for 2020. The development demand is then compared with the environmental sensibility to develop a Growth Allocation procedure.