An Algorithm for Deploying a Crime Directed (Tactical) Patrol Force

This paper presents an algorithm for deploying a crime directed patrol force. The optimization problem is formulated as the allocation of N patrol among R high crime regions so as to maximize the weighted probability of a patrol intercept of a crime. The algorithm has full sensitivity analysis capabilities. This capability is critical because the model's input parameters include (1) crime weights, which in general will have a subjective component, (2) crime descriptive data, which are difficult to estimate, and (3) crime frequency data, which are likely to change with time. The paper presents an illustrative application of the algorithm. The resultant allocation is compared to a strategy which allocates patrol units in direct proportion to each region's total crime rate. The optimal allocation significantly increased the probability of an intercept.