To be or not to be? A spatial predictive crime model for Rochester
share
This project uses a spatial model (Geographically Weighted Regression) to relate various physical and social features to crime rates. Besides making interesting predictions from basic data statistics, the trained model can be used to predict on the test data. The high accuracy of this prediction on test data then allows us to make predictions of crime probabilities in different areas based on the location, the population, the property rate, the time of the day/year and so on. This then further gives us the idea that an application can be built to help people traveling around Rochester be aware when and if they enter crime prone area.
READ FULL TEXT