基於GA-BP 神經網絡的恐怖主義犯罪態勢研究
王娟、白芝禕、管雨翔

摘要：
對涉恐案件進行犯罪態勢預測研究有助於幫助維護國內外的安全與穩定，尤其是對“一帶一路”沿線國家的恐怖主義犯罪態勢預測，將對建立“一帶一路”沿線國家安全預警與防範機制有重要作用。基於GTD全球恐怖主義數據庫2016年至2021年之間的中亞、南亞、東亞、東南亞四個地區的數據，以神經網絡算法為基礎，採用GA遺傳算法和BP神經網絡模型相互融合並優化的算法，對未來所處的安全形勢預測評估。“一帶一路”沿線地區的犯罪態勢逐年下降。南亞地區的涉恐案件最為突出，其犯罪數量高居不下主要源於A、B兩國國內環境等多方面因素。爆炸式襲擊是涉恐案件的主要犯罪手段，需要各國聯合對武器、爆炸類原材料等嚴格管理。

關鍵詞：
BP神經網絡 遺傳算法 涉恐案件 犯罪態勢預測

Prediction of Terrorism-Related Crime Situation Based on GA-BP Neural Network
Wang Juan; Bai Zhiyi; Guan Yuxiang

Abstract：
The research on crime situation prediction of terrorism-related crimes is helpful to maintain security and stability at home and abroad. In particular, the prediction of terrorism-related crime situation of the countries along “the Belt and Road” will play an important role in establishing the security early warning and prevention mechanisms in those regions. By utilizing the data of Central Asia, South Asia, East Asia and Southeast Asia from 2016 to 2021 in GTD global terrorism database, based on neural network algorithm, which integrates and optimizes the GA genetic algorithm and BP neural network model to predict and evaluate the future security situations. The crime situation in the areas along the way of “the Belt and Road” decreased year by year. Terrorism-related crimes are the most prominent in South Asia, and the high number of crimes is mainly due to the domestic environment of countries A and B, among other factors. Explosive attacks are the main means of terrorism-related crimes, necessitating collaborative efforts for the strict management of weapons, explosive raw materials and so on.

Key words：
BP Neural Network Genetic Algorithm Terrorism-related Crimes Crime Situation Prediction

DOI：
https://doi.org/10.63334/esfsmjournal.i7.2025.03.01