Eid Mohamed Hamdy, Kamel Mohamed Sayed, Amer Anwar Sayed Kamel, Szűcs Péter
Institute of Environmental Management, Faculty of Earth Science, University of Miskolc, 3515, Miskolc, Egyetemváros, Hungary.
Geology Department, Faculty of Science, Beni-Suef University, Beni-Suef, 65211, Egypt.
Heliyon. 2024 Oct 4;10(19):e38684. doi: 10.1016/j.heliyon.2024.e38684. eCollection 2024 Oct 15.
Armed conflicts, as significant human phenomena, profoundly impact populations and reflect a state's capacity to fulfill its responsibilities. These conflicts arise from various causes, necessitating robust predictive models to understand their spatial distribution. This study employs the Bivariate Frequency Ratio (FR) method to spatially predict the occurrence of armed conflicts across the East African States, drawing on 42 political geography-related criteria. The development of the predictive model involved classifying the region into five conflict-prone categories influenced by critical political geography factors. Geospatial datasets, curated in a GIS environment, were sourced from approved online portals. The findings indicate that Burundi exhibits the highest vulnerability to armed conflict, followed closely by Rwanda, Uganda, and Somalia. Ethiopia and South Sudan show a moderate risk, while predictions for Zimbabwe, Zambia, and Mozambique suggest lower likelihoods of conflict. The model's accuracy was validated using the Receiver Operating Characteristic (ROC) curve, demonstrating its effectiveness. Furthermore, the model's applicability extends to other regions, offering a valuable tool for global conflict prediction.
武装冲突作为重大的人类现象，对民众产生深远影响，并反映出一个国家履行其责任的能力。这些冲突由多种原因引发，因此需要强大的预测模型来了解其空间分布。本研究采用双变量频率比（FR）方法，利用42个与政治地理相关的标准，对东非各国武装冲突的发生情况进行空间预测。预测模型的开发涉及根据关键政治地理因素将该地区划分为五个易发生冲突的类别。在地理信息系统（GIS）环境中整理的地理空间数据集来自经批准的在线门户网站。研究结果表明，布隆迪对武装冲突的脆弱性最高，紧随其后的是卢旺达、乌干达和索马里。埃塞俄比亚和南苏丹显示出中等风险，而对津巴布韦、赞比亚和莫桑比克的预测表明冲突可能性较低。该模型的准确性通过接收者操作特征（ROC）曲线进行了验证，证明了其有效性。此外该模型的适用性扩展到其他地区，为全球冲突预测提供了一个有价值的工具。
