University of Bath

Predicting terrorist events: Opportunities & challenges

In this talk, Dr Andre Python will briefly review recent literature on prediction of terrorist events, including a few important papers in armed conflict.

26 Mar 20191.15pm
to
26 Mar 20192.05pm
Free
Snapshot of a weekly prediction of terrorist attacks
Snapshot of a weekly prediction (week 600) of terrorist attacks in Iraq, Iran, Afghanistan, Pakistan using XGBoost machine learning algorithm.

Terrorism persists as a worldwide threat, as exemplified by the ongoing lethal attacks perpetrated by Islamic State in Iraq and Syria, Al Qaeda in Yemen, and Boko Haram in Nigeria. In response, states deploy various counterterrorism policies, the costs of which could be reduced through efficient preventive measures. Predictive models that can account for complex spatiotemporal dependencies have not yet been investigated, despite their potential for providing guidance to prevent terrorism.

In this talk, Dr Andre Python will first briefly review recent literature discussing prediction of terrorist events, including few important papers in armed conflict. He will then introduce the preliminary results of a study which uses a machine-learning algorithm to predict the locations of terrorist events a week ahead at fine spatial scale in Iraq, Iran, Afghanistan, and Pakistan. He will conclude the talk by discussing the opportunities and challenges in predicting terrorist events.