Chapter 3(2.8)
Chapter 7Multivariate techniques with textParallel embedded
system design lab7.1 IntroductionData collecting by taking
measurements often unpredictableMeasurement errorRandomly selected
objectsMultivariate statisticsTechniques for analyzing
simultaneously text miningPrincipal components analysis 7.2 Basic
statisticsSample variance
Z Scores Applied to PoeComputing how a data value compares to a
data setConverting value dimensionless
7.2 Basic statistics
7.2 Basic statistics
7.2 Basic statistics
Z-score of 4for The Forest Reverie7.2.2 Word correlations among
Poes short storiesZ-score has problem about comparing between units
Correlation
7.2 Basic statistics
7.2 Basic statistics
7.2 Basic statistics
7.2 Basic statistics
7.2 Basic statistics
7.2 Basic statistics7.2.3 Correlations and cosines
Computing cosine using matrix multiplication7.2 Basic
statistics7.2.3 Correlations and cosines
7.2 Basic statistics7.2.4 Correlations and
covariancesCovariance
Correlation
7.3 Basic linear algebraSquare matrix X, M (having at least one
nonzero vector)Satisfying
n by n correlation and covariance matrix n real, orthogonal
eigenvectors with n real eigenvalues
= number7.3 Basic linear algebra7.3.1 2 by 2 correlation
matrices
7.3.1 2 by 2 correlation matrices
7.3 Basic linear algebra
7.3.1 2 by 2 correlation matrices
7.3 Basic linear algebra
7.3.1 2 by 2 correlation matrices
7.3 Basic linear algebra
First : linear function of the originalSecond : vector C has
unit lengthThird : each pair of and (i j)Four : the variances of ,
, , are ordered from largest to smallest7.4 Principal components
analysis
7.4.1 Finding the principal componentsCorrelation
matrixz-scoreCovariance matrixoriginal data values7.4 Principal
components analysis7.4 Principal components analysis
Computing the principal components with function procmp()
7.4 Principal components analysis
Using summary() on the output of prcomp()
7.4 Principal components analysis
Computing the principal components using the covariance
matrix7.4 Principal components analysis
Another PCA example with Poes short stories7.4 Principal
components analysis
Another PCA example with Poes short stories
7.4 Principal components analysis
7.4 Principal components analysis7.4.4 RotationsOnly changing
orientation but not the shape of any objectAny rotation in
n-dimensions is representable by an n-by-n matrix
A PCA preserves all of the information
