| 基于贝叶斯网络的海盗袭击事件影响因素 |
| |
| 引用本文： | 蒋美芝,吕靖,王爽.基于贝叶斯网络的海盗袭击事件影响因素[J].中国航海,2019(2):87-92. |
| |
| 作者姓名： | 蒋美芝 吕靖 王爽 |
| |
| 作者单位： | 大连海事大学交通运输工程学院 |
| |
| 基金项目： | 国家自然科学基金(71473023);中央高校基本科研业务费专项资金(3132016359);教育部人文社会科学规划基金(16YJAZH030) |
| |
| 摘 要： | 为分析海盗袭击事件的影响因素,减少海盗袭击事件的发生,保障海上运输安全,引入贝叶斯网络理论,利用期望最大化(Expectation Maximization,EM)算法和联合树推理算法构建海盗袭击事件分析模型。根据东南亚海域海盗袭击事件的实际数据获得贝叶斯网络(Bayesian Networks, BN)结构各节点之间的概率依赖关系,并对海盗袭击成功节点进行敏感性分析。分析结果表明:外界环境、船舶自身风险和时间是影响海盗袭击事件发生的主要因素。
|
| 关 键 词： | 水路运输 海运安全 贝叶斯网络 敏感性分析 海盗袭击 期望最大化算法 联合树算法 |
Influencing Factors of Piracy Attacks with Bayesian Network |
| |
| JIANG Meizhi,LYU Jing,WANG Shuang.Influencing Factors of Piracy Attacks with Bayesian Network[J].Navigation of China,2019(2):87-92. |
| |
| Authors: | JIANG Meizhi LYU Jing WANG Shuang |
| |
| Institution: | (College of Transportation Engineering,Dalian Maritime University,Dalian 116026,China) |
| |
| Abstract: | The Bayesian network theory is introduced to analyze the influencing factors of the pirate attacks. The analytic model of piracy attacks is established. The EM(Expectation Maximization) algorithm and the joint tree algorithm are used for determining the structure parameters and conducting sensitivity analysis. Case study for Southeast Asian sea area shows that those nodes of external environment, ship’s own risk and time are of the most significances with regard to the influences on the probability nodes of piracy attacks. |
| |
| Keywords: | waterway transportation maritime safety Bayesian network sensitivity analysis piracy attack EM algorithm joint tree algorithm |
| 本文献已被 CNKI 维普 等数据库收录！ |
