Principal Component Analysis for Histogram-Valued Data

来源 :2010年第十届中日统计研讨会 | 被引量 : 0次 | 上传用户:tjpu0510420215
下载到本地 , 更方便阅读
声明 : 本文档内容版权归属内容提供方 , 如果您对本文有版权争议 , 可与客服联系进行内容授权或下架
论文部分内容阅读
  In recent year,datasets have become increasingly larger and more complex in a variety of fields with the development of the Internet and the improvement of computer performance.It is difficult to analyze them by using classical statistical methods and the new methods are required.Then,Symbolic data analysis (in short SDA) has been attracting much attention and have been studied as the one of useful methods for analyzing large and complex datasets.SDA is defined as an extended data analysis for more complex data (table) called "symbolic data (table) ".Typical symbolic data are the interval-valued data and the histogram-valued data.Here,we deal with the histogram-valued data and consider a principle component analysis (in short PCA) for the histogram-valued data.When we aggregate individuals into classes or have multi values for a single observation,the histogram-valued data are obtained.
其他文献
  There are not so many studies on Japanese economic data analyzed by threshold model or double threshold GARCH model (here after DT-GARCH model).Then in this
会议
  In traditional financial time series area,the closing price is always the base of research on shares.It happens that the loss of some observation causes the
  Around the world,financial risk pre-warning is feasible and imperative.This paper is based on the principle of selection indicator system,and designs a stro
会议
  Volatility of financial assets prices is the important component in asset pricing and financial risk management,and jump is the discontinuous parts in the v
  In this paper,we study an influence of malaria in China.Malaria is epidemic in the tropical and the temperate zones,and moreover Plasmodium falciparum has h
会议
  The paper employs the Bayesian approach to establish acceptance sampling plans for life tests with interval censoring.Assume that interval data have a multi
会议
  Against the backdrop of the progress of globalization and accompanying social and economic transitions,there has been in Japan of late an increase in the gr
会议
  This work stems from the need for statistical shape analysis of open curves in 2D or 3D space.By the abstractions such as ignoring local thicknesses and ext
会议
  We consider the delimitation of the distances of spatial point pattern interactions in a Neyman-Scott spatial cluster field and note its geometrical implica
会议
  In Murakami and Tanida (2009),we demonstrated that spreading peoples distrust in Japanese public pension system through human network would have big effect
会议