Chapter 12:
Principal Component Analysis
12.1 Introduction
12.2 Geometric and algebraic bases of principal components
12.4 Plotting of principal components
12.5 Principal components from the correlation matrix
12.6 Decide how many components to retai

Chapter 8:
Discriminant Analysis:
Description of Group
Separation
8.1 Introduction
8.2 Discriminant Function for Two Groups
8.4 Discriminant Analysis for Several Groups
8.5 Standardized Discriminant Function
8.6 Tests of Significance
8.1 Introduction

Chapter 4:
The Multivariate Normal
Distribution
4.1 Univariate and multivariate normal densities
4.2 Properties of multivariate normal random variables
4.3 Estimation in the multivariate normal
Univariate Normal Distribution
A random variable y N (, 2)

Chapter 11:
Canonical Correlation
11.1 Introduction
11.2 Canonical Correlations and Canonical Variates
11.3 Properties of Canonical Correlations
11.4 Tests of Significance
11.1 Introduction
p
Multiple correlation R = R2 measures the linear relationsh

Chapter 4:
The Multivariate Normal
Distribution
4.1 Univariate and multivariate normal densities
4.2 Properties of multivariate normal random variables
4.3 Estimation in the multivariate normal
Univariate Normal Distribution
A random variable y N (, 2)

Chapter 9:
Classification Analysis:
Allocation of Observations to
Groups
9.1 Introduction
9.2 Classification into two groups
9.3 Classification into several groups
9.4 Estimating misclassification rates
9.5 Improved estimates of error rates
9.7 Nonp

Chapter 5:
Tests on One or Two Mean
Vectors
5.1 Multivariate versus univariate tests
5.2 Tests on with known
5.3 Tests on when is unknown
5.4 Comparing two mean vectors
5.7 Paired observations test
5.1 Multivariate vs Univariate Tests
Suppose y = (y

Chapter 6:
Multivariate Analysis of
Variance (MANOVA)
6.1 One-way models
6.2 Comparison of four MANOVA tests
6.3 Contrasts
6.5 Two-way classification
We begin with a review of univariate analysis of variance (ANOVA) before covering
multivariate
analys

5-11First, we known this is a T-test on multivariate statistics with an unknown variance matrix.
'
H 0 : =( 6 11 )
As the problem has indicated that the Hotellings T2-Test Null-Hypothesis is :
2
2
ThereforeWhen y 0 and s are replaced by y 0 and S, we obta