Scribd
Upload
8 views5 pages

Quiz Week 11

Uploaded by

Rama Bhushan
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
8 views5 pages

Quiz Week 11

Uploaded by

Rama Bhushan
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/ 5

Consider the following case study and answer the questions below.

You study 100


individuals along 10 demographic features. Your experiment requires PCA analysis of
the given data. Answer the following questions. The options mention intervals: choose
the one in which the correct value lies.

Q1: How many maximum PCs can be generated?

a) 0-3
b) 4-7
c) 7-10
d) 11-14
e) None of the above.

Hint: The lower dimension is of demographic features (i.e., variables), which is 10.
Hence, a maximum of10 PCs can be generated.

Q2: The following eigenvalues are obtained.

PC1 PC2 PC3 PC4 PC5 PC6 PC7 PC8 PC9 PC10
3.5 1.5 1 1 0.5 0.5 0.5 0.5 0.5 0.5

What is the total variation explained by PC1?

a) 0-10%
b) 10-20%
c) 20-30%
d) 30-40%
e) None of the above

Hint: 3.5+1.5+1+1+0.5*6=10, PC1 variation explained =3.5/10 =35%


Q3: The following eigenvalues are obtained.

PC1 PC2 PC3 PC4 PC5 PC6 PC7 PC8 PC9 PC10
3.5 1.5 1 1 0.5 0.5 0.5 0.5 0.5 0.5

Compute the sum of squared distance, SSD for PC1 (d=distance of projectile point from
origin)

a) 0-10
b) 10-20
c) 20-30
d) 30-40
e) None of the above

Hint: Eigen value = SSD /(n-1), SSD = Eigen value*(n-1) =3.5*9=31.5

Q4: The following eigenvalues are obtained.

PC1 PC2 PC3 PC4 PC5 PC6 PC7 PC8 PC9 PC10
3.5 1.5 1 1 0.5 0.5 0.5 0.5 0.5 0.5

What is the total variation explained by PC2.

a) 0-10%
b) 10-20%
c) 20-30%
d) 30-40%
e) None of the above

Hint: 3.5+1.5+1+1+0.5*6=10, PC2 variation explained =1.5/10 =15%


Q5: The following eigenvalues are obtained.

PC1 PC2 PC3 PC4 PC5 PC6 PC7 PC8 PC9 PC10
3.5 1.5 1 1 0.5 0.5 0.5 0.5 0.5 0.5

Compute the sum of squared distance, SSD for PC2 (d=distance of projectile point from
origin)

a) 0-10
b) 10-20
c) 20-30
d) 30-40
e) None of the above

Hint: Eigen value = SSD /(n-1), SSD = Eigen value*(n-1) =1.5*9=13.5

Q6: The following eigenvalues are obtained.

PC1 PC2 PC3 PC4 PC5 PC6 PC7 PC8 PC9 PC10
3.5 1.5 1 1 0.5 0.5 0.5 0.5 0.5 0.5

If one chooses 8 PCs, PC1 to PC8, how much variation in the data can be explained.

a) 55-65%
b) 65-75%
c) 75-85%
d) 85-95%
e) None of the above

Hint: 3.5+1.5+1+1+0.5*6=10, PC1-8 variation explained =9/10 =90%

Q7: You decide to choose only two PCs. The first PC, PC1, has a slope of 1/5 on XY plane
(X= Income, Y= Expenditure), what is the slope of PC2.

a) -6 to -4
b) -4 to -2
c) 2 to 4
d) 4 to 6
e) None of the above
Hint: Since the second PC is orthogonal to the first PC, the slope of PC1*slope of PC2 =
-1. Therefore, its slope would be -5.

Q8: Suppose instead of 10 variables, you started with only 2 variables (X=income and
Y=expenditure) and found a slope of ½ for PC1. What would be the loadings of income
on PC1?

a) 0.0-0.3
b) 0.3-0.6
c) 0.6-0.9
d) 0.9-1.2
e) None of the above

Hint: Loading of income = 2/ sqrt(1^2+2^2) = 2/sqrt(5)= 0.8944

Q9: Suppose instead of 10 variables, you started with only 2 variables (X=income and
Y=expenditure) and found a slope of ½ for PC1. What would be the loadings of
expenditure on PC1?

a) 0.0-0.3
b) 0.3-0.6
c) 0.6-0.9
d) 0.9-1.2
e) None of the above

Hint: Loading of income = 1/ sqrt(1^2+2^2) = 1/sqrt(5)= 0.4472

Q10: Suppose instead of 10 variables, you started with only 2 variables (X=income and
Y=expenditure) and found a slope of ½ for PC1. What would be the slope of PC2?

a) -3 to -1
b) -1 to 1
c) 1 to 3
d) 3 to 5
e) None of the above

Hint: Since the second PC is orthogonal to the first PC, the slope of PC1*slope of PC2 =
-1. Therefore, the slope of PC2= -2

You might also like