LSTM在国家间双边关系预测中的应用  

Self-Attention LSTMs for Cooperation-Conflict Prediction Methods

作　　者：练良珏 胡梅婷 邹昆 LIAN Liangjue;HU Meiting;ZOU Kun(College of Software,Nanchang Hangkong University,Nanchang,China,330063)

机构地区：[1]南昌航空大学软件学院,南昌330063

出　　处：《福建电脑》2024年第6期64-67,共4页Journal of Fujian Computer

摘　　要：为预测国家之间双边关系的演化趋势,本文构建了一个国家之间“合作-冲突”的预测模型。该模型基于自注意力机制的长短期神经网络,以国家事件动作为特征,根据最近发生的一系列事件,预测下一个时间单元的冲突事件和合作事件的数量并以戈尔德斯坦量评分进行衡量。通过公开的数据集验证模型的预测效果,结果表明本文模型具有较好的预测效果与国家关系判断能力。To predict the evolution trend of bilateral relations between countries,this article constructs a prediction model of"cooperation conflict"between countries.The model is based on a long-term and short-term neural network with self attention mechanism,characterized by national event actions.Based on a series of recent events,it predicts the number of conflict and cooperation events in the next time unit and measures them using the Goldstein scale score.The predictive performance of the model was validated through publicly available datasets,and the results showed that the proposed model has good predictive performance and ability to judge national relations.

关 键 词：国家间关系 冲突与合作 预测模型 长短期神经网络 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程]