基于聚类分析算法的舰船数据挖掘  

Clustering analysis algorithm for ship data mining

作　　者：张传美[1] ZHANG Chuan-mei(Science and Technology College of NCHU,Nanchang 330034,China)

机构地区：[1]南昌航空大学科技学院,江西南昌330034

出　　处：《舰船科学技术》2020年第6期166-168,共3页Ship Science and Technology

基　　金：江西省教育科学“十三五”规划2018年度一般课题资助项目(18YB408)

摘　　要：目前应用的舰船数据挖掘准确度较低,因此设计一种聚类分析算法的海上舰船数据挖掘。在数据挖掘中应用聚类分析算法,需要对数据进行预处理,得到文本数据特征向量集,利用模糊集体现近似关系,根据隶属度的取值将模糊聚类的思想演变为目标函数,将数据集按照目标函数,划分为具有较小差距的群组,初步得到数据挖掘结果,经过评估后,筛选出准确度最高的数据,作为数据挖掘的最终结果。至此完成了聚类分析算法的海上舰船数据挖掘的研究。通过实验表明,设计的数据挖掘准确度平均为92%,比传统的数据挖掘准确度高17.2%,验证了设计的聚类分析算法的海上舰船数据挖掘在提高挖掘准确度方面的可靠性。The accuracy of ship data mining is low, so a clustering analysis algorithm is designed for ship data mining at sea. Application of clustering analysis algorithm in data mining, need for data preprocessing, text feature vector data set,fuzzy collective approximate relationship now, according to the membership value of the fuzzy clustering thought evolved as objective function, the data set according to the objective function, divided into groups, with smaller gap preliminary data mining results, after evaluation, select the highest accuracy of data, as a result of data mining. So far, the research on data mining of Marine ships based on clustering analysis algorithm has been completed. The experimental results show that the designed data mining accuracy is 92% on average, 17.2% higher than the traditional data mining accuracy, which verifies the reliability of the designed clustering analysis algorithm in improving the mining accuracy of offshore ships.

关 键 词：聚类算法 数据挖掘 VSM模型 

分 类 号：U674.7[交通运输工程—船舶及航道工程]