How Deep Neural Networks Help in Prediction of Future Terrorist...
Govciooutlook

How Deep Neural Networks Help in Prediction of Future Terrorist Activities

By Gov CIO Outlook | Tuesday, February 23, 2021

Machine learning approaches have recently been investigated to develop counter-terrorism strategies based on Artificial Intelligence (AI). Since deep learning has lately gained more popularity in the field of Machine Learning (ML), these techniques are explored to recognize the behavior of terrorist activities.

FREMONT, CA: Terror remains one of the most critical challenges to the civilization of today. Terrorism disrupts law and order conditions in society and impacts the quality of human life and makes them physically and mentally repressed and deprives them of life enjoyment. The more civilizations have developed, the more people will explore various mechanisms to protect humanity from terrorism. Different anti-terrorism strategies have been used to safeguard people's lives in society and enhance living quality in general.

Machine learning approaches have recently been investigated to develop counter-terrorism strategies based on Artificial Intelligence (AI). Since deep learning has lately gained more popularity in the field of Machine Learning (ML), these techniques are explored to recognize the behavior of terrorist activities.

Five separate models based on a Deep Neural Network (DNN) are developed to explain the actions of terrorist attacks, such as whether the attack is going to be successful or not? Or is the attack going to be suicidal or not? Or what kind of weapon is going to be used in the attack? And what sort of assault is going to take place? Or what country is going to be attacked for?

The models are implemented in a single-layer Neural Network (NN), a 5-layer DNN, and three standard machine learning algorithms: logistic regression, Support Vector Machine (SVM) Naive Bayes. The efficiency of the DNN is contrasted with the NN and the three machine learning algorithms, and it shows that the performance of the DNN is more than 95 percent in terms of accuracy, precision, recall, and F1-Score. In contrast, the ANN and conventional machine learning algorithms have achieved a maximum accuracy of 83 percent. This instance concludes that DNN is a useful model for forecasting the actions of terrorist acts.

Weekly Brief

Read Also