Scribd
Upload
373 views

Math4424: Homework 4: Deadline: Nov. 21, 2012

This document contains instructions for homework problems involving principal component analysis. Problem 8.1 asks to find the principal components and proportion of variance explained for a given covariance matrix. Problem 8.3 asks to find the principal components and comments on eigenvectors for a different covariance matrix. Problem 8.4 asks to find principal components and proportion of variance explained for a covariance matrix involving correlations.

Uploaded by

Moshi Zeri
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
373 views

Math4424: Homework 4: Deadline: Nov. 21, 2012

This document contains instructions for homework problems involving principal component analysis. Problem 8.1 asks to find the principal components and proportion of variance explained for a given covariance matrix. Problem 8.3 asks to find the principal components and comments on eigenvectors for a different covariance matrix. Problem 8.4 asks to find principal components and proportion of variance explained for a covariance matrix involving correlations.

Uploaded by

Moshi Zeri
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 2

Math4424: Homework 4

Deadline: Nov. 21, 2012

8.1. Determine the population principal components Y1 and Y2 for the covariance
matrix
 
5 2
Σ= .
2 2

Also, calculate the proportion of the total population variance explained by the first
principal component. 
8.3. Let
 
2 0 0
Σ= 0 4 0 
0 0 4

Determine the principal components Y1 , Y2 and Y3 . What can you say about the
eigenvectors (and principal components) associated with eigenvalues that are not
distinct? 
8.4. Find the principal components and the proportion of the total population
variance explained by each when the covariance matrix is
 2 
σ σ2ρ 0
1 1
Σ =  σ2ρ σ2 σ2ρ  , −√ < ρ < √ . 
0 σ2ρ σ2 2 2

8.5. (a). Find the eigenvalues of the correlation matrix


 
1 ρ ρ
ρ =  ρ 1 ρ ,
ρ ρ 1

Are your results consistent with (8-16) and (8-17)?

1
(b). Verify the eigenvalue-eigenvector pairs for the p×p matrix ρ given in (8-15).
8.6. Data on x1 = sales and x2 = profits for the 10 largest companies in the
world were listed in Exercise 1.4 of Chapter 1. From Example 4.12,
   
155.60 7476.45 303.62
x̄ = , S= .
14.70 303.62 26.19

(a). Determine the sample principal components and their variances for these
data. (You may need the quadratic formula to solve for the eigenvalues of S).
(b). Find the proportion of the total sample variance explained by ŷ1 .
(c). Sketch the constant density ellipse (x − x̄)0 S−1 (x − x̄) = 1.4, and indicate
the principal components ŷ1 and ŷ2 on your graph.
(d). Compute the correlation coefficients rŷ1 ,xk , k = 1, 2. What interpretation, if
any, can you give to the first principal component? 

You might also like