SOM算法在海洋大数据挖掘中的应用初探
详细信息 查看官网全文
- 英文论文题名：Exploratory research on SOM algorithm in mass oceanic data excavation
- 论文作者：孟春霞 ; 李桂娟 ; 车树伟 ; 王东梅 ; 白金
- 英文论文作者：MENG Chun-xia ; LI Gui-juan ; CHE Shu-wei ; WANG Dong-mei ; BAI Jin ; Science and Technology on Underwater Test and Control Laboratory
- 年：2016
- 作者机构：水下测控技术国防科技重点实验室;
- 论文关键词：自组织映射方法 ; 大数据 ; 特征挖掘 ; 统计分析
- 英文论文关键词：Self-Organizing Map ; mass data ; characteristic excavation ; statistical analysis
- 会议召开时间：2016-10-28
- 会议录名称：2016年全国声学学术会议论文集
- 语种：中文
- 分类号：P733.2;TP311.13
- 学会代码：OGSM
- 会议名称：2016年全国声学学术会议
- 会议地点：中国湖北武汉
- 主办单位：中国声学学会
- 学会名称：中国声学学会
- 页数：3
- 文件大小：473k
- 原文格式：D
- 会议级别：全国
摘要
自组织映射方法是一种基于竞争学习的无监管神经网络,它能把高维的输入数据映射到低维空间并且保留数据的拓扑结构,特别适合于解决数据的聚类问题。海洋环境噪声是海洋中的背景声场,海洋环境噪声的声级大小受附近行船、工业的人为活动、海洋生物或风雨等多种因素的影响。本文研究了自组织映射的基本原理和算法;利用SOM方法得到了20Hz-31.5k Hz宽频带海洋环境噪声特性的统计变化规律,并初步分析了海洋环境噪声级与行船和风速的内在关系。本文的研究结果为从海量数据中挖掘样本之间的关联关系提供技术参考。
The Self-Organizing Map(SOM) is an unsupervised neural network based on competitive learning, and can solve the problem that the center of clustering is unknown. The ambient noise is the background sound field, which sound pressure levels depend on nearby ship, various human activities and marine organisms, passing rain showers and so on. SOM's theory and the implementation of algorithm are studied in this paper. The statistical variability of broadband ambient noise at frequencies between 20 Hz and 31.5 k Hz is obtained using SOM, and the preliminary analysis of the relationship between ambient noise level and vessels is carried out. The results provide the technical reference to obtain relevant relationship between input samples from mass data.
The Self-Organizing Map(SOM) is an unsupervised neural network based on competitive learning, and can solve the problem that the center of clustering is unknown. The ambient noise is the background sound field, which sound pressure levels depend on nearby ship, various human activities and marine organisms, passing rain showers and so on. SOM's theory and the implementation of algorithm are studied in this paper. The statistical variability of broadband ambient noise at frequencies between 20 Hz and 31.5 k Hz is obtained using SOM, and the preliminary analysis of the relationship between ambient noise level and vessels is carried out. The results provide the technical reference to obtain relevant relationship between input samples from mass data.
引文
[1]Liu,Y.,and R.H.Weisberg,A review of self-organizing map applications in meteorology and oceanography[J].Self-Organizing Maps.Applications and Novel Algorithm Design,2011,253–272.
[2]Nikolina Rako,Mapping underwater sound noise and assessing its sources by using a self-organizing maps method[J].J.Acoust.Soc.Am.,2013;133(3):1368-1376.
[3]Nikolina Rako,Marta Picciulin.Spatial and temporal variability of sea ambient noise as an anthropogenic pressure index:the case of Vres-Losinj archipelago,Croatia[J].Journal of the Marine Biological Association of the United Kingdom,2013;93(1):27-36.
[2]Nikolina Rako,Mapping underwater sound noise and assessing its sources by using a self-organizing maps method[J].J.Acoust.Soc.Am.,2013;133(3):1368-1376.
[3]Nikolina Rako,Marta Picciulin.Spatial and temporal variability of sea ambient noise as an anthropogenic pressure index:the case of Vres-Losinj archipelago,Croatia[J].Journal of the Marine Biological Association of the United Kingdom,2013;93(1):27-36.