 PCA is ________?

Marked Answer :

Unsupervised

Correct Answer :

Unsupervised

Total Marks :1MARKS OBTAINED1

 Why is PCA needed in the Data Science field?

Marked Answer :

In Dimensionality Reduction

Correct Answer :

In Dimensionality Reduction

Total Marks :1MARKS OBTAINED1

 Extract only features and scale the data using StandardScaler() , compute min covariance?

Marked Answer :

-0.57

Correct Answer :

-0.57

Total Marks :1MARKS OBTAINED1

 Extract only features and scale the data using StandardScaler() , how many eigen_values
will be calculated?

Marked Answer :

64

Correct Answer :

64

Total Marks :1MARKS OBTAINED1

 from the above eigenvalues and eigenvectors, create eigenpair and calculate maximum
cumulative explained variance?

Marked Answer :
100

Correct Answer :

100

Total Marks :1MARKS OBTAINED1

 Why is cumulative explained variance required before applying PCA?

Marked Answer :

To check the maximum components to be selected

Correct Answer :

To check the maximum components to be selected

Total Marks :1MARKS OBTAINED1

 from the below plot, select the range of components can be selected?

Marked Answer :

30-40

Correct Answer :

10-20

Total Marks :1MARKS OBTAINED0

 Fit the data to PCA and compute the maximum explained_variance_ratio?

0.12

Correct Answer :

0.12

Total Marks :1MARKS OBTAINED1

 Which of the following is the correct syntax for finding the eigenvalues and eigenvectors
with give covariance matric (cov_mat)?

Marked Answer :

eig_vals, eig_vecs =np.linalg.eig(cov_mat)

Correct Answer :

eig_vals, eig_vecs =np.linalg.eig(cov_mat)

Total Marks :1MARKS OBTAINED1

 Choose the correct order of steps to choose principal components:

A . compute the covariance matrix.

B. choose the principal components from Eigen values and Vectors.

C. collecting the data.

D. compute Eigen values and vectors.

E. standardization.

Marked Answer :

C->E->A->D->B

Correct Answer :

C->E->A->D->B

Total Marks :1MARKS OBTAINED1

 What correction is required in the following lines of code to create covariance matrix?

scaler = StandardScaler()

standardized_data = scaler.fit_transform(data)

covariance_matrix = np.cov(standardized_data)

Marked Answer :
Need to perform transpose of standardized_data

Correct Answer :

Need to perform transpose of standardized_data

Total Marks :1MARKS OBTAINED1

 What exactly do you understand by the given code?

pca = PCA(n_components=0.95)

Marked Answer :

We are looking for the principal components that cumulatively explain 95% of the variance in the
data.

Correct Answer :

We are looking for the principal components that cumulatively explain 95% of the variance in the
data.

Total Marks :1MARKS OBTAINED1

 How is the first principal component determined in PCA?

Marked Answer :

It is the component with the most variance.

Correct Answer :

It is the component with the most variance.

Total Marks :1MARKS OBTAINED1

 Which of the following non-zero vector stays parallel after matrix multiplication

Marked Answer :

EigenVector

Correct Answer :

EigenVector

Total Marks :1MARKS OBTAINED1

 The Output of PCA is always a new representation of data with a lower dimension than
the original data representation.

Marked Answer :
TRUE

Correct Answer :

TRUE

Total Marks :1MARKS OBTAINED1

 What is the difference between PCA and Linear Regression?

Marked Answer :

Euclidean distance calculated in PCA, Vertical distance calculated in Linear Regression.

Correct Answer :

Euclidean distance calculated in PCA, Vertical distance calculated in Linear Regression.

Total Marks :1MARKS OBTAINED1

 How to Determine the number of principal components to retain in PCA?

Marked Answer :

Cumulative variance plot

Correct Answer :

Cumulative variance plot

Total Marks :1MARKS OBTAINED1

 What is the minimum and maximum limit of n_components value we can give in PCA?

Marked Answer :

Minimum 2, Maximum 100% of independent columns count

Correct Answer :

Minimum 2, Maximum 100% of independent columns count

Total Marks :1MARKS OBTAINED1

 Use the Gym.csv dataset and Apply PCA in the given dataset, what is the minimum
n_components required to attain 0.99% of variance?

Marked Answer :

Correct Answer :
4

Total Marks :1MARKS OBTAINED0

 Apply PCA in the given dataset(gym.csv) with n_components as 2, Then get the original
(approx.) data back from PCA, what is the difference in sum of variance of all columns
before and after applying PCA?

Marked Answer :

5.85

Correct Answer :

16.26

Total Marks :1MARKS OBTAINED0

 Is it appropriate to use PCA as a method for preventing overfitting in machine learning

Marked Answer :

FALSE

Correct Answer :

FALSE

Total Marks :1MARKS OBTAINED1

 What is the sum of eigenvalues calculated for the first 8 records(gym.csv) with the
independent features?

Marked Answer :

98.32

Correct Answer :

101.76

Total Marks :1MARKS OBTAINED0

 Compare the performance of two Linear Regression models(gym.csv):

1. The first model uses the raw data without any preprocessing.

2. The second model applies PCA with n_components set to 4 before fitting the data.

Both models should be generated using train_test_split with test_size=0.2 and random_state=24
during the train-test split.
Your task is to calculate the Root Mean Squared Error (RMSE) for both models and determine
whether the performance of the second model (PCA-based) is better compared to the first model
(raw data).
Marked Answer :

Model 2 performance is better than Model 1

Correct Answer :

Model 2 performance is not better than Model 1

Total Marks :1MARKS OBTAINED0

 Compute min correlation in the given dataset?

Marked Answer :

-0.0592

Correct Answer :

-0.0592

Total Marks :1MARKS OBTAINED1

 Compare the performance of two Linear Regression models(gym.csv):

1. The first model uses the raw data without any preprocessing.

2. The second model applies standardscaler only to independent features and then
apply PCA with n_components set to 4 before fitting the data.

Both models should be generated using train_test_split with test_size=0.2 and random_state=24
during the train-test split.
Your task is to calculate the Root Mean Squared Error (RMSE) for both models and determine
whether the performance of the second model (StandardScalar & PCA-based) is better compared
to the first model (raw data).

Marked Answer :

Model 2 performance is not better than Model 1

Correct Answer :

No change in performance

Total Marks :1MARKS OBTAINED0

 While Selecting Principal Components, the eigenvalues are arranged in descending order,
indicating the amount of variance explained by each component(gym.csv)?

Marked Answer :

TRUE

Correct Answer :
TRUE