Interpretable machine learning predicts terrorism worldwide | TechXplore

Dubo/Shutterstock.com

An international research team led by Dr. Andre Python of Zhejiang University’s Center of Data Science investigates machine learning models adept at predicting and explaining at a fine spatiotemporal level the eventuality of terrorist attacks committed by non-state actors from outside legit warfare (non-state terrorism) around the globe, as published in Science Advances. While machine learning algorithms perform well in locations where terrorism is prevalent, predicting occurrences in areas where terrorism has not been prevalent for a long time remains difficult. Even at the fine spatial and temporal resolution, algorithms may demonstrate a rather high overall accuracy. 

Source: https://techxplore.com/news/2021-08-machine-terrorism-worldwide.html

RELATED ARTICLES

Subscribe to our newsletter