Innovative Thinking
Choosing its spatial targets based on a new and innovative clustering algorithm, the Daily Crime Forecast then capitalizes on the spatial and temporal distribution of crime by effectively assessing the relevance of attributes that include location, time of day, day of the week, day of the month and month within incident data. Using several temporal and spatial factors together, a better and more accurate prediction can be made to pin-point times and locations where crime is likely to occur.
Effective Clustering
Instead of providing 'hot spots' areas based on user-chosen parameters as in most spatial hot spotting routines, the Daily Crime Forecast builds its hot spots based on available resources. This new paradigm ensures that the target areas and times make use of all available personnel and prioritizes hot spots in order to maximize coverage.
In the Field
The Daily Crime Forecast has been deployed to a live environment within the Edmonton Transit System where Peace Officers have willingly adopted the model to guide their patrols. Managers and supervisors also use the Daily Crime Forecast to readily identify resource allocation priorities within the transit system.
Comments received about the Daily Crime Forecast:
"Not so much as fishing with a fish finder but having the fish jump into the boat. Worked like a charm."
Michael Kostek
Peace Officer, Edmonton Transit System
Andrew Nicoll
Peace Officer, Edmonton Transit System
Superior Performance
In a series of tests that compared the predictive performance of the Daily Crime Forecast model to conventional hot spot techniques, the Daily Crime Forecast (DCF) consistently outperformed the spatial/temporal hotspot.
Fact Sheet
Want more information? Click the link below to download our fact sheet and get more descriptions and details.