Algorithm uses real-time data analysis to predict illegal activity

Algorithm uses real-time data analysis to predict illegal activity

A team from the University of Surrey, UK, and Georgia Institute of Technology, USA, has developed an algorithm which can predict illegal activity using real-time data analysis.

In a paper published in Computational Statistics and Data Analysis, the team details a new way to potentially predict illegal activity, utilising a method similar to that used in weather forecasting which incorporates real-time analysis of urban crime data.

The ability of the new algorithm to incorporate real-time data analysis can supplement Epidemic Type Aftershock Sequence (ETAS) models, which are the systems most often applied to urban crime data for the purposes of predicting crime. ETAS models create a grid-based map that has been able to predict twice as much crime as a single dedicated analyst.

How does the algorithm work?

The value of the new algorithm, dubbed the Ensemble Poisson Kalman Filter (EnPKF), is that it is able to provide real-time forecasts for the crime rate and predict the possibility of a crime repeating in a certain area based on the input of data in real time. It can also predict where short-term crime hotspots could arise, allowing law enforcement authorities to better prioritise the allocation of their resources.

The researchers also believe that the algorithm could make effective and more accurate forecasts if applied to models other than ETAS, and be used to predict everything from train delays to earthquake aftershocks. It has already been tested against data on more than 1000 violent gang crimes in Los Angeles, USA, between 1999 and 2002.

What has the research team said about its findings?

Dr David Lloyd, who led the study from the University of Surrey’s Department of Mathematics, expressed optimism about the potential applications of the new algorithm, and the next round of testing, which seeks to make the EnPKF stronger.

He said: “We are cautiously excited about the Ensemble Poisson Kalman Filter, an approach that has given us an insight into when crime can be predicted, and has shown us the importance of using real-time data to make the overall system stronger… It is important to remember that EnPKF, and algorithms similar to this, are tools used to help our law enforcement who work hard to keep our communities safe.”

  • LinkedIn
  • Twitter
  • Facebook

LEAVE A REPLY

Please enter your comment!
Please enter your name here