A Toy Model Study for Long-Term Terror Event Time Series Prediction with CNN
- PDF / 1,430,094 Bytes
- 21 Pages / 439.37 x 666.142 pts Page_size
- 12 Downloads / 164 Views
A Toy Model Study for Long‑Term Terror Event Time Series Prediction with CNN Aishvarya Kumar Jain1 · Christian Grumber1 · Patrick Gelhausen1 · Ivo Häring1 · Alexander Stolz1 Received: 23 July 2019 / Accepted: 4 December 2019 © Springer Nature Switzerland AG 2019
Abstract Followed by the 9/11 attacks in 2001 and the subsequent events, terrorism and other asymmetrical threat situations became increasingly important for security-related efforts of most western societies. In a similar period, the development of data gathering and analysis techniques especially using the methods of machine learning has made rapid progress. Aiming to utilize this development, this paper employs artificial neural networks for long-term time series prediction of terrorist event data. A major focus of the paper lies on the specific use of convolutional neural networks (CNNs) for this task, as well as the comparison to the performance of classical methods for (long-term) time series prediction. As the databases like Global Terrorism Database and Fraunhofer’s terrorist event database are not extensive enough to train a deep learning method, a simple toy model for the generation of time series data from one or more terrorist groups with defined properties is established. Metrics for comparison of the different approaches are collected and discussed, and a customized sliding-window metric is introduced. The study shows the principle applicability of CNNs for this task and offers constraints as well as possible extensions for future studies. Based on these results, continuation and further extension of data collection efforts and ML optimization techniques are encouraged. Keywords Multi-step · Long term · Time series prediction · Artificial neural network · Convolutional network · Terrorism
* Aishvarya Kumar Jain [email protected] 1
Fraunhofer Institute for High‑Speed Dynamics, Ernst-Mach-Institut, Am Klingelberg 1, 79588 Efringen‑Kirchen, Germany
13
Vol.:(0123456789)
A. K. Jain et al.
1 Introduction The dynamics of conflicts have changed during the course of the last century. While the security situation, especially in western societies, traditionally was dominated by scenarios of potential or actual symmetrical warfare, modern conflicts often evolve around asymmetric situations. This development has been emphasized and further accelerated by the infamous 9/11 attacks and the subsequent events. Consequently, terrorism has become a topic of increasing importance for considerations on governmental, economic and societal levels. A major characteristic of terror is its unpredictability. To unfold its negative effects, it needs to generate a ubiquitous danger to health and material assets. This yields a strategy that actively tries to hide patterns or early warnings. Naturally to prevent or mitigate the impact, identifying such patterns, i.e., via time series analysis, ever since has become a field of high interest for science and authorities. A huge variety of instruments in classical data science and statistics
Data Loading...