%0 Journal Article
%T Effectiveness of selected AI models in predicting victims of terrorist attacks
%A Tuszyński, Jakub
%J Terroryzm – studia, analizy, prewencja
%V 2023
%R 10.4467/27204383TER.23.035.18337
%N Numer 4 (4)
%P 471-502
%K AI, uczenie maszynowe, terroryzm, ofiary, klasyfikacja
%@ 2720-4383
%D 2023
%U https://ejournals.eu/czasopismo/terroryzm/artykul/effectiveness-of-selected-ai-models-in-predicting-victims-of-terrorist-attacks
%X 
	Terrorism continues to be a problem for many countries around the world. The article compares the effectiveness of selected machine learning algorithms in predicting the victims of terrorist attacks in order to answer the question of whether they can serve as one of the anti-terrorist tools. An exploratory data analysis was carried out, and selected trends and characteristics of terrorist attacks were discussed. Some measures for evaluating the classification algorithms used in the study are presented, and potential directions for further research are indicated.

	*The article is based on a master’s thesis entitled Effectiveness of selected AI models i predicting victims of terrorist attacks, defended at the Faculty of Journalism, Information and Bibliology, University of Warsaw. The author used excerpts from chapters 3 and 6. The thesis was awarded in the 12th edition of the competition of the Head of the Internal Security Agency for the best doctoral, master’s or bachelor’s thesis concerning state security in the context of intelligence, terrorist, economic threats.