In the program’s first three weeks, the model has proven to be 71 percent accurate in predicting the place and day where crimes have occurred, said Deputy Chief Steve Clark. In other words, 71 percent of the time the model told an officer to be at a location, a crime was in progress or was reported.
At least one person has been arrested because of a patrol check initiated by the system, he said.
Police recently gave eight years of crime reports for residential burglaries, vehicle burglaries and vehicle thefts to an applied mathematics professor at Santa Clara University. The system essentially maps the time, location and recurrence of past crimes to help police predict crime and tailor their patrols.–Santa Cruz police have success with predictive policing
For this building snowmobiles post, I wanted to draw upon a new crime fighting technology and explore the idea of it’s possible uses. The idea here is to use predictive analysis, much like what retailers use for product research or what researchers use for earthquake prediction, and use it to predict where crime is most likely to happen to get resources to efficiently cover those areas via patrols.
So the question I ask is if this actually works for crime, then why not apply it to warfighting? And especially COIN and today’s conflicts, where the war is long and there is time to collect statistics of attacks and instances that would be needed to build such a model? Or how about for anti-piracy or for the drug war down in Mexico? The key is if you have statics over the course of several years, then a model could be made. And if war planners are wanting to use their resources more efficiently because they have less forces to use, or the host nation is limited in resources, then predictive warfighting might help with the more efficient use of manpower on the battlefield.
If anything, much like with policing, it will be the guy on the ground who patrols their areas daily that will have the intuition of where to go and how best to cover their AO’s. But what about units that cycle in and out of the battlefields? Where is their intuition coming from if they have never been to that AO? So predictive analysis might help in the transition periods and help build that intuition of the new forces. This predictive analysis will also make it easier to make judgements about setting up patrols. You could combine human intuition/experience/orientation with this predictive analysis, and make a better plan of operations.
Predictive policing also helps the COIN forces by efficiently guiding the local police forces to areas they need to be. With places like Afghanistan, you might have officers who do not want to go in certain areas or dwell more in certain areas than they should, or are not trained enough to recognize patterns, or they come from other parts of the country. They too could benefit from this predictive analysis to further reinforce their intuition. But it could also help determine if that police force is working efficiently.
The fear though is depending upon this predictive analysis entirely. To me it is an interesting tool that needs to be tested more to see where it can be most effective, or where it could fit in to the overall strategy for crime fighting or warfighting. Interesting stuff and definitely check out all of the articles and information posted below if you would like to read more about it.
On a final note, Santa Cruz and other police departments throughout the nation looked at this new system as a way to more efficiently use their police forces to deal with crime. But they were also looking at it because of economic reasons because there is less money available to fund police departments these days. So more and more departments will be looking at cost cutting measures, while still being able to ‘protect and serve’ their communities. So what say you? –Matt
#60: Fighting Crime With Mathematics
12.16.2010
By Daniel Lametti
One major problem in crime-fighting is that a police crackdown in one neighborhood may simply push criminal behavior into a nearby area. In March two mathematicians, working with an anthropologist and a criminologist, announced a way to quantify this reaction (pdf).
“Crimes tend to cluster together in space and time, forming hot spots,” says UCLA mathematician Martin Short, the study’s lead author. Drawing on real-world data, his team developed a model showing that hot spots come in two varieties. One type forms when an area experiences a large-scale crime increase, such as when a park is overrun by drug dealers. Another develops when a small number of criminals—say, a pair of burglars—go on a localized crime spree.
The model suggests that a focused police response can relatively easily extinguish larger hot spots because the criminals there scatter randomly, making it unlikely that they will resume coordinated unlawful activity nearby. But for smaller crime waves, crooks just migrate together into an adjacent neighborhood, where they are likely to start another spree. By analyzing police reports as they come in, Short hopes to determine which type of hot spot is forming so police can handle it more effectively.
Link to Discover article here.
—————————————————————-
UC MaSC Project
Funded by the Human Social Dynamics Program at NSF, the UC MaSC Project centers on theoretical, methodological and empirical work to develop analytical and computational models of crime pattern formation. Crime mapping forms a key feature of current approaches to understanding offender behavior and is a tool used increasingly by police departments and policy makers for strategic crime prevention. However, despite the availability of sophisticated digital mapping and analysis tools there is a substantial gap in our understanding of how low-level behaviors of offenders lead to aggregate crime patterns such as crime hot spots. Thus, for example, we are unable to specify exactly why directed police action at crime hot spots sometimes leads to displacement of crime in space but, surprisingly, often can also lead to hot spot dissipation and a real reduction in crime incidences.Agent-based modeling offers a potential avenue for developing a quantitative understanding of crime hot spot formation built from the bottom-up around offender behavior. Agent-based models are not only more consistent with the scale of decisions that offenders actually take, but they also open the door to the development of custom statistics that are designed to answer specific behavioral questions less tractable in general statistical models. However, there is also concern that agent-based simulations can lead to erroneous results either because of poor model design or errors in model implementation that go undetected. A solution to this problem is to design simulations around well-studied analytical models where the model behavior can be tested against sound analytical expectations. Only following such testing should simulation models be extended into areas that cannot be treated analytically and, only subsequent to this, into applied contexts.?The UC MaSC Project has four components. 1. Drawing on methods in statistical physics and the mathematics of swarms, we are developing formal models of offender movement and target selection in variously structured environments.
2. We plan to extend these baseline models to consider offender behavior on abstract urban street networks.
3. We will then integrate both model types with Geographic Information Systems (GIS) by exploring the spatial properties of simulated crime maps.
4. At each stage of model development, empirical tests will be conducted against spatial crime data provided by the Los Angeles, San Diego and Long Beach police departments. We will concentrate empirical testing on comparing simulated crime prevention interventions with known changes in urban planning and policing strategies within these southern Californian cities.
Simulated offender movement and abstract mapping of crime locations. (a) Simulated offender following a Lévy mobility strategy. Lévy mobility shows clusters of short distance flights interspersed with longer distance flights. (b) Mapped crime locations assuming that criminal opportunities are uniformly distributed in space and that offenses occur only at the end points of Lévy flights. (c) Inverse distance weighted interpolation of offense locations showing crime hot spots generated by the underlying Lévy mobility strategy.?The UC MaSC Project will help clarify the quantitative relationships between criminal behavior, criminal opportunities and policing and may provide insight into how to design better crime prevention strategies, contributing to a broader dialog on homeland security. Simultaneous development of mathematical and simulation models, as well as continuous empirical testing, will provide a guide for the experimental use of these tools in the social sciences, while the broad interdisciplinary foundation of the project will provide a model for collaboration between mathematicians and social scientists. The educational component, including planned supplemental REU training, will provide an excellent venue for developing the research careers of students and postdoctoral associates at all levels.
Simulation of spontaneous burglary hotspot nucleation. Panels on the left show the cumulative distribution of locations visited in a two-dimensional plane. Panels on the right show the cumulative distribution of burgled houses. Simulation code written by Jon Azose, a Harvey Mudd undergraduate Math student participating in the NSF Harvey Mudd-UCLA Research Training Group program run by Andy Bernoff.
Link to UC MaSC here.
—————————————————————-Santa Cruz police have success with predictive policing
By STEPHEN BAXTER
07/18/2011
A new predictive policing method that started in July has added to officers’ intuition of where and when to target patrols in the city.
Santa Cruz police are the first in the nation to implement a predictive method – which is gaining interest nationally – and leaders are already seeing some positive changes.
In the program’s first three weeks, the model has proven to be 71 percent accurate in predicting the place and day where crimes have occurred, said Deputy Chief Steve Clark. In other words, 71 percent of the time the model told an officer to be at a location, a crime was in progress or was reported.
At least one person has been arrested because of a patrol check initiated by the system, he said.
Police recently gave eight years of crime reports for residential burglaries, vehicle burglaries and vehicle thefts to an applied mathematics professor at Santa Clara University. The system essentially maps the time, location and recurrence of past crimes to help police predict crime and tailor their patrols.
Starting July 1, police administrators have asked the program to produce a list of the highest percentage time and place of crime that day.
When officers on a shift are not responding to calls for service, they are asked to check on those areas for potential crimes of auto or home burglaries, said police spokesman and crime analyst Zach Friend.
Officers might normally check those problem areas by intuition, but police said the new system adds some statistical backup.
“Many of the locations we suspected would have a high probability, but this adds a verified math component to it. This enhances the officer’s intuition,” said Deputy Chief Steve Clark said. “We can more efficiently target these locations.”
Santa Cruz police typically discuss the predicted hot spots with officers during roll call meetings before each shift, Friend said. Officers receive a map and list of hours and places with corresponding probabilities of crime.
Police aren’t disclosing the crime hot spots in the city in an effort to thwart criminals, but a full evaluation of the program’s first six months is expected in December or January 2012. The Sentinel first reported on the system in January.
Police presence could also deter crime in areas they check, so police will examine whether the crime rate drops in areas that they target, Friend said.
Clark said the technology might seem strange at first, but he compared it to radios in police cars. That was an idea that at its inception met with skepticism is now considered indispensable, he said.
“You have to be willing to look to the future and use these tools,” Clark said. “It’s going to help us target the precious few resources that we have.”
Story here.
—————————————————————
Crime Analyst’s Blog here.New York Times story here.