Hostname: page-component-7b7567c8c-z4zkx Total loading time: 0 Render date: 2026-11-17T20:05:17.388Z Has data issue: false hasContentIssue false

Advanced Forecasting of Military-Targeted Terrorism Using Time Series Analysis

Published online by Cambridge University Press:  23 March 2026

Heejun Shin*
Affiliation:
SCH Disaster Medicine Center, Soonchunhyang University Bucheon Hospital, Republic of Korea
Rights & Permissions [Opens in a new window]

Abstract

Core share and HTML view are not available for this content. However, as you have access to this content, a full PDF is available via the 'Save PDF' action button.
Introduction:

Military-targeted terrorism poses a persistent threat to global security, impacting national defense and international stability. Accurate forecasting of such incidents is crucial for developing effective counter-terrorism strategies. This study leverages advanced time series analysis to predict future trends in military-targeted terrorism incidents, providing insights for policymakers and military strategists.

Methods:

This study utilizes historical data from the Global Terrorism Database (GTD) from 1970 to 2020 to forecast incidents from 2021 to 2030. Three predictive models, such as ARIMA, Random Forest, and Long Short-Term Memory (LSTM) networks, were employed to capture patterns and predict future incidents. Model performance was evaluated using Mean Absolute Error (MAE), Root Mean Square Error (RMSE), and R-squared (R²) metrics. The LSTM model was further optimized through hyperparameter tuning to enhance prediction accuracy.

Results:

The LSTM model demonstrated superior performance compared to ARIMA and Random Forest models. Specifically, the LSTM achieved an MAE of 10.2 and an RMSE of 12.5, with an R² value of 0.82, indicating a strong fit and predictive capability. The ARIMA model recorded an MAE of 12.3 and RMSE of 14.8, while the Random Forest model had an MAE of 15.7 and RMSE of 18.2. The forecasts revealed an increasing trend in military-targeted terrorism incidents over the next decade, with predicted peaks of 65 incidents in 2022 and 71 incidents in 2023.

Conclusion:

Integrating LSTM networks with traditional time series models enhances the predictive accuracy of military-targeted terrorism forecasts. These insights enable policymakers to allocate resources effectively, plan strategic interventions, and improve preparedness against potential threats. Future research should incorporate additional data sources, such as social media analytics and geopolitical factors, to refine predictions and enhance situational awareness. This study underscores the importance of leveraging advanced machine learning techniques to address complex security challenges.

Information

Type
Meeting Abstracts
Copyright
© The Author(s), 2026. Published by Cambridge University Press on behalf of World Association for Disaster and Emergency Medicine