This is the ninth of 10 chapters on "Spatial Modeling II" from the user manual for CrimeStat IV, a spatial statistics package that can analyze crime incident location data.
This chapter, "Time Series Forecasting," presents time series methods useful for tactical deployment of police resources. These methods make two determinations relevant to the allocation of police resources. First, they determine what crime levels are expected in police zones, patrol districts, census tracts, or other areas of a jurisdiction, given past time trends. Second, they determine whether any new crime patterns (large increases or decreases) are emerging in the jurisdiction. The first determination is made by using extrapolative time series models to forecast expected crime levels by geographic area. The second determination uses observed departures from the expected crime levels as the basis for detecting new crime patterns. This chapter presents standard models and methods that have long been used in industry for time series forecasting and detection, making them available in highly optimized and automated computer code tailored for crime analysis. The chapter first presents overviews of time series data and extrapolative forecast methods. It then presents details on exponential smoothing models, which are among the simplest but most accurate forecasting methods, along with classical decomposition for estimation of seasonal adjustments. The chapter concludes with a discussion of the early detection of time series patterns, first as a concept and then as implemented in CrimeStat. Exponential smoothing forecasts are the basis for detection, so that all the methods covered in this chapter work together. 7 figures and 14 references
Downloads
Related Datasets
Similar Publications
- Place-based Correlates of Motor Vehicle Theft and Recovery Measuring Spatial Influence Across Neighbourhood Context
- Overview of Crime Travel Demand Modeling (CrimeStat IV: A Spatial Statistics Program for the Analysis of Crime Incident Locations, Version 4.0)
- Predicting dynamical crime distribution from environmental and social influences