Methods for interpreting a self-organized map in data analysis.pdfVIP

下载本文档

2
0
约1.23万字
约 6页
2017-04-13 发布于江苏
举报
版权申诉

Methods for interpreting a self-organized map in data analysis.pdf

1、原创力文档（book118）网站文档一经付费（服务费），不意味着购买了该文档的版权，仅供个人/单位学习、研究之用，不得用于商业用途，未经授权，严禁复制、发行、汇编、翻译或者网络传播等，侵权必究。。
2、本站所有内容均由合作方或网友上传，本站不对文档的完整性、权威性及其观点立场正确性做任何保证或承诺！文档内容仅供研究参考，付费前请自行鉴别。如您付费，意味着您自己接受本站规则且自行承担风险，本站不退款、不进行额外附加服务；查看《如何避免下载的几个坑》。如果您已付费下载过本站文档，您可以点击这里二次下载。
3、如文档侵犯商业秘密、侵犯著作权、侵犯人身权等，请点击“版权申诉”（推荐），也可以打举报电话：400-050-0827(电话支持时间：9:00-18:30)。
4、该文档为VIP文档，如果想要下载，成为VIP会员后，下载免费。
5、成为VIP后，下载本文档将扣除1次下载权益。下载后，不支持退款、换文档。如有疑问请联系我们。
6、成为VIP后，您将拥有八大权益，权益包括：VIP文档下载权益、阅读免打扰、文档格式转换、高级专利检索、专属身份标志、高级客服、多端互通、版权登记。
7、VIP文档为合作方或网友上传，每下载1次，网站将根据用户上传文档的质量评分、类型等，对文档贡献者给予高额补贴、流量扶持。如果你也想贡献VIP文档。上传文档

Methods for interpreting a self-organized map in data analysis

Methods for Interpreting a Self-OrganizedMap in Data AnalysisSamuel Kaski, Janne Nikkila, and Teuvo KohonenHelsinki University of Technology,Neural Networks Research Centre,P.O. Box 2200, FIN-02015 HUT, FINLANDAbstract. The Self-Organizing Map (SOM) can be used for formingoverviews of multivariate data sets and for visualizing them on graphicalmap displays. Each map location represents certain kinds of data itemsand the value of a variable in the representations can be visualized inthe corresponding locations on the map display. Such component planedisplays contain all the information needed for interpreting the map butinformation about the relations of the variables remains implicit. Wehave developed methods that visualize explicitly the contribution of eachvariable in the organization of the map at dierent locations. It is alsopossible to measure the contribution of each variable in the cluster struc-ture within an area of the map to summarize, for instance, the charac-teristics of clusters.1. IntroductionThe SOM algorithm [2, 3] forms a mapping of a usually two-dimensional maplattice into the high-dimensional data space. There is a model vector connectedto each point of the discrete lattice. The model vectors are situated in the dataspace; they act as an ordered set of models of dierent types of data items.The map can be used as an ordered groundwork for illustrating dierentaspects of the data set. In addition to visualizing the values of the originalvariables as component planes (examples are shown in Fig. 2a) the map can beused to visualize the clustering tendency of the data in dierent regions of thedata space. The model vectors follow the distribution of the data items andtherefore the distances between the model vectors connected to neighboringpoints on the map lattice are shorter in clustered areas than in sparser regions.The so-called U-matrix display [4], an example of which is shown in Fig. 1,depicts the distances between neighboring model vectors a