基于粗糙集数据约简的海事事故致因研究

Rough set reduction-based study on the causes of marine accidents

为使结果更具适应性和普遍性，运用粗糙集数据处理系统对由6个因素组成的海船船员适任决策表进行数据约简．结果表明：较低航海教育程度者对海事发生的隶属度较高；船长对海事发生的隶属度最高；具有12个月以上本职务服务资历的海船船员，其服务资历与发生海事的隶属度成反比．对一般及以下等级海事事故，值班时间在0004～0008的海船船员对海事事故发生的隶属度较高；34～39年龄段的海船船员具有较高的隶属度．对大事故及以上等级海事事故，值班时间在0000～0004的海船船员对海事事故的隶属度较高；海船船员年龄与海事发生的隶属度成反比．

To make the results more adaptable and universal, data is reduced for seafarers competency decision table composed of six factors by using rough set exploration system. Results show that less educated seafarers have higher rough membership degrees, and the master has the highest membership degree. The membership degree is at an inverse ratio to seafarers with 12 months and more sea experience. For minor or lower marine accidents, the seafarers between 34 and 39 years old have the highest membership degrees. High membership degree in watch-keeping time is 0004-0008. For big and higher accidents, high membership degree in watch-keeping time is 0000--0004, and the membership degree is an inverse ratio to seafarer＇ s ages.