What is PredPol?
The PredPol system utilizes a machine-learning algorithm to generate predictions of potential criminal activity. The algorithm is trained on historical event data from the local police department, typically spanning 2 to 5 years, for each new city or jurisdiction where the system is deployed. As new incident reports are received daily from the department's records management system, the algorithm is updated to incorporate the latest information and refine its predictive capabilities. This data-driven approach allows the system to adapt to the evolving crime patterns and trends within a specific geographic area
Highlights
- Employs a machine-learning algorithm to generate predictive crime analytics
- Relies on historical crime data (2-5 years) to train the prediction model
- Integrates with the department's records management system to continuously update the algorithm
- Provides location-specific predictive insights tailored to each deployment area