Using Neural Network to Predict Unmanned Aerial Vehicle Strikes in Pakistan

Unmanned Aerial Vehicles (UAVs) are pilotless aircrafts which were originally introduced for military purposes. They play a significant role in US war-on-terrorism. Since 2004, US has attacked many targets in Pakistan using UAVs. Pakistan condemns these attacks and denies the allegation of their hidden approval. In this context, we propose a neural network based model to minimize the adverse effects of these illegal attacks by predicting their frequency along with the number of militant killed, civilian causalities and number of injuries. The predictive model is trained using Pakistan UAV strikes data and results show that the proposed model predicts these variables with good accuracy and small RMSE (Root Mean Square Error).