Predicting future crime has left the realm of the futuristic science fiction and is now part of the very day reality, at least for the officers of the LAPD.  These officers are now using the power of big data to determine hotspots for crime and as a result lowering crime across the Los Angeles metropolitan area.  What started with a mathematical program to predict earthquakes has now contributed to a reduction of 33% in burglaries, 21% in violent crime, and 12% in property crime across the area where the algorithms are being applied.

Feeding the mathematical model developed by Assistant Professor George Moher with huge amounts of past crime data – 13 million crimes spanning the last eight decades – since criminal activity often has similar ‘aftershocks’ as earthquakes, big data now allows the LAPD to predict the areas that will be hotspots of crime in the near future.  In collaboration with the University of California and PredPol, whose goal it is to “place officers at the right time and location to give them the best chance of preventing crime” the algorithm and results were further fine tuned.

When the pilot project was launched, police officers were hesitant to start using the program and resistant to the idea that a data program would better tell them where and how to do their job than years of experience on the force.  But application of the program and monitoring the results in real time has swayed opinions – as have the stellar results mentioned earlier.  Now, the model is also being updated in real time with crime data as it comes in, to hone the predictive power of the big data even more.

Finally, the power of big data analytics does not end with predicting crime.  Big data is now also being used to fight insurance fraud and putting the crime that occurs into the context of the larger picture in which it is taking place.  Correlations can now also be found, as Shaun Hipgrave, a former police officer and current security consultant for IBM, illustrates:  “When you use big data you can see the relationships between one family and another troubled family and you see the absences from school.”  The stellar results across the board from this trial program has now resulted in the PredPol software being rolled out for trials in over 150 American cities, and is likely to expand and spread further across the country.


Read more here

(Image Credit:  jondoeforty1)

Previous post

Data Science Retreat Teaching About Big Data in Berlin

Next post

Teradata Teams with Siemens