3. Assult Risk Prediction in Chicago, IL
A new survey shows that nearly half of Chicago residents feel “very unsafe” in the city as a whole, and less than a quarter of Chicagoans feel safe in the city where they live. In addition, less than 30 percent of residents feel safe in their neighborhoods, while the same survey 2021 fall showed that 45 percent of the public felt safe in their neighborhoods.
A geospatial risk model is a regression model. The dependent variable is the occurrence of discrete events like crime, fires, etc. Predictions from these models are interpreted as ‘the forecasted risk/opportunity of that event occurring here’.
So I want to find out the regular of assault number that happens on the road. If we use a geo-spatial risk model, would it be possible for the police station to place more policers in places where road assault is likely to occur. Can such a model reduce the occurrence of road assault?
In order to avoid other assault information, like assault happens in the apartment or buildings. The dataset only contains assault happens on the street, sidewalk, and alleys.
Data Visualization
Model Evaluation
