Application of cluster analysis of risk profiles and health consequences of maritime terrorist act

INTRODUCTION. Maritime terrorism poses a particular threat to international security, complicated by the high likelihood of medical consequences and limited medical response capabilities in port and shipboard environments. Despite the existence of descriptive studies, there have been no studies to date that utilize multivariate analysis and clustering methods to identify typical scenarios that take into account sanitary and irreversible losses. OBJECTIVE. To identify stable clusters of maritime terrorist attacks with non-zero losses based on a set of method-means-object characteristics and characterize their destructive potential for the purposes of disaster medicine. MATERIALS AND METHODS. The study was conducted using the GTD (Global Terrorism Database), which includes 209,707 terrorist attacks from 1970 to 2020. Of these, 69,772 terrorist attacks were identified, and 35,591 attacks with ≥ 1 fatality were selected using keywords. The analysis used normalization, logarithmic loss transformation, one-hot encoding of categorical features, and standardization. Clustering was performed using the squared means method, and the optimal number of clusters was selected based on the silhouette coefficient, Kalinski–Harabas, and Davis–Bouldin indices. The results were interpreted using PCA visualizations and distribution profiles. RESULTS. Three stable clusters were identified. The first (high-risk) is characterized by the predominance of explosives, a focus on private individuals, military personnel, and police, and the highest health consequences (median: 9 killed and 15 injured per event; percentile (p) 90–32 killed and 52 injured). The second cluster includes predominantly low-yield explosive attacks; the third includes armed attacks on security forces with moderate consequences. The peak of activity was recorded in 2013–2016, when the proportion of high-risk incidents reached its maximum values. DICUSSION. The obtained results confirm the heterogeneity of maritime terrorism and allow us to identify three qualitatively different threat regimes. Medical consequences range from isolated injuries to mass casualties requiring triage and evacuation. The cluster approach provides a basis for developing medical response scenarios, with an emphasis on preparedness for catastrophic events involving explosive hazards. CONCLUSION. The proposed methodological approach expands the analytical framework for maritime terrorism research, enabling a transition from descriptive characteristics to a reproducible classification of terrorist attacks by risk profile. This creates a practical basis for planning medical support, allocating resources, and improving the regulatory framework in maritime security.

1. LaFree G., Dugan L. Introducing the Global Terrorism Database. Terrorism and Political Violence, 2007, Vol. 19, No. 2, pp. 181–204.

2. START. Global Terrorism Database (GTD) Codebook: Methodology and Variables. University of Maryland, 2022, 147 p.

3. LaFree G., Yang S. M. The Impact of Global Trends on Terrorism. International Journal of Comparative and Applied Criminal Justice, 2010, Vol. 34, No. 1, pp. 1–19.

4. Bueger C. What is Maritime Security? Marine Policy, 2015, Vol. 53, pp. 159–164.

5. Lehr P. (ed.). Violence at Sea: Piracy in the Age of Global Terrorism. London: Routledge; 2006, 287 p.

6. Евдокимов В.И. Медико-биологические последствия терроризма в мире: монография / Всероссийский центр экстренной и радиационной медицины им. А. М. Никифорова МЧС России, Санкт-Петербургский медико-социальный институт. СПб.: Измайловский, 2024. 101 с. (Серия «Чрезвычайные ситуации в мире и России»; вып. 3) [Evdokimov V. I., Medical and biological consequences of terrorism in the world: a monograph / A. M. Nikiforov All-Russian Center for Emergency and Radiation Medicine of the Ministry of Emergency Situations of Russia, St. Petersburg Medical and Social Institute. St. Petersburg: Izmailovsky, 2024. 101 p. (Series “Emergencies in the world and Russia”; issue 3) (In Russ.)].

7. Евдокимов В. И. Терроризм и его медико-биологические последствия в мире (2011-2020 гг.) // Медико-биологические и социально-психологические проблемы безопасности в чрезвычайных ситуациях. 2024. № 1. С. 14–33 [Evdokimov V. I., Terrorism and its medical and biological consequences in the world (2011-2020). Med.-biol. and social psychology. security probation in Russia. situations, 2024, No. 1, pp. 14–33 (In Russ.)]. doi: 10.25016/2541-7487-2024-0-1-14-33.

8. Шуленин Н. С., Мавренков Э. М., Шуленин К. С., Киселев В. С. Морской терроризм в цифрах: анализ терактов, угроз и перспектив противодействия // Морская медицина. 2025. Т. 11, № 2. С. 120–134 [Shulenin N. S., Mavrenkov E. M., Shulenin K. S., Kiselyov V. S. Maritime terrorism in figures: analysis of terrorist attacks, threats and prospects of counteraction. Marine medicine, 2025, Vol. 11, No. 2, pp. 120–134 (In Russ.)]. doi: 10.22328/2413-5747-2025-11-2-120-134. EDN XIQEPQ.

9. Шуленин Н. С., Лемешкин Р. Н., Мавренков Э. М., Шуленин С. Н. Структура и последствия морских террористических актов: данные для оценки рисков и планирования медицинского реагирования // Морская медицина. 2025. Т. 11, № 3. С. 93–110 [Shulenin N. S., Lemeshkin R. N., Mavrenkov E. M., Shulenin S. N. Structure and consequences of maritime terrorist acts: data for risk assessment and medical response planning. Marine medicine, 2025, Vol. 11, No. 3, pp. 93–110 (In Russ.)]. doi: https://dx.doi.org/10.22328/2413-5747-2025-11-3-93-110; EDN: https://eLibrary.ru/KAELVA.

10. Murphy M. Contemporary Piracy and Maritime Terrorism: The Threat to International Security. London: Routledge; 2007, 224 p.

11. Klein N. Maritime Security and the Law of the Sea. Oxford: Oxford University Press; 2011, 372 p.

12. Raymond C. Z. Maritime Terrorism in Southeast Asia: A Risk Assessment. Terrorism and Political Violence, 2009, Vol. 21, No. 2, pp. 274–293.

13. Kaufman L., Rousseeuw P.J. Finding Groups in Data: An Introduction to Cluster Analysis. New York: Wiley; 2005, 368 p.

14. MacQueen J. Some Methods for Classification and Analysis of Multivariate Observations. Proceedings of the Fifth Berkeley Symposium on Mathematical Statistics and Probability, 1967, Vol. 1, pp. 281–297.

15. Rousseeuw P. J. Silhouettes: A Graphical Aid to the Interpretation and Validation of Cluster Analysis. Journal of Computational and Applied Mathematics, 1987, Vol. 20, pp. 53–65.

16. Calinski T., Harabasz J. A Dendrite Method for Cluster Analysis. Communications in Statistics, 1974, Vol. 3, No. 1, pp. 1–27.

17. Davies D. L., Bouldin D. W. A Cluster Separation Measure. IEEE Transactions on Pattern Analysis and Machine Intelligence, 1979, Vol. PAMI-1, No. 2, pp. 224–227.

18. Jolliffe I. T., Cadima J. Principal Component Analysis: A Review and Recent Developments. Philosophical Transactions of the Royal Society A: Mathematical, Physical and Engineering Sciences, 2016, Vol. 374, № 2065, pp. 20150202.

19. Pedregosa F., Varoquaux G., Gramfort A., Michel V., Thirion B., Grisel O., Blondel M., Prettenhofer P., Weiss R., Dubourg V., Vanderplas J., Passos A., Cournapeau D., Brucher M., Perrot M., Duchesnay É. Scikit-learn: Machine Learning in Python. Journal of Machine Learning Research, 2011, Vol. 12, pp. 2825–2830.