您现在的位置: > 环球 > 人文海外 > 正文

基于多元统计图的高维数据投降维方法及运用切

2019-06-16 05:47  来源:原创   字号:T | T

  燕地脊父亲学工学硕士学位论文

  Abstract

  After the and when atmulti—class

  analyzingchallengedifficultyfacing

  dimensional methodof and

  hierarchical

  high data,the multi—layer

  reductionbasedonmultivariatechart hasbeen

  dimensionality representation

  inthis

  proposedpaper.

  and in reductionare

  Firstly,theimportantimpendencedimensionality

  the ofrelationalliteratures.Itisthevitalandmutual

  analyzedby investigation

  thathow of

  reduces dimensionaldataandfind

  problem dimensionalityhigh

  themethodtounderstandandsolvethe in science

  problemeasilymany fields.

  The distribution,statistical of

  geometrical high

  dimensionaldata in definitionof

  type mathematics,the

  been

  reductionandintrinsic have discussedindetail.

  dimeusionality

  reductionmethodsof dimensionaldata

  Secondly,thedimensionality high

  into

  areclassifiedtwo withdifferent linearmethods

  categories projection.The

  which areinvolvedwith

  usually Component

  Principal

  LinearDiscriminant introduced nonlinear

  Analysis(LDA),etc.isbfiefly.The

  methodswhose includeMulti·Dimensional

  generally Scale(MDS),

  地脊东方修盖父亲学硕士学位论文

  ModernUrban ResearchBasedon

  Landscape Regional

  Characteristics

  …TakeJinan’s asan

  Special Example

  Landscape

  Tian theartIearns)

  Fei(Designing

  DirectedProf.Zhou

  by Changii

  ABSTRACT

  Withthe of and andtheconstuctionofcities

  developmenteconomysociety,urbanization

  havemade afewsmalltownshave into

  greatprogress.Quiet developedbigmetropolises.

  ofurbanconstructioncan’t forthe

  However,thegreat stand favorable

更多