Introduction To Machine Learning IIT MADRAS

Introduction To Machine Learning

IMP QUESTIONS PART 3

Question 17:
In a neural network, what does the term "epoch" refer to?
A) The number of layers in the network
B) The learning rate of the network

AM
C) One complete pass through the entire training dataset
D) The activation function used in the network the training and validation
datasets
Accepted Answer: C) One complete pass through the entire training dataset

Question 18:
AY
Which evaluation metric is suitable for imbalanced classification problems?
SW
A) Accuracy
B) Precision
C) F1-Score
D) Mean Squared Error
Accepted Answer: C) F1-Score
Y
M

Question 19:
What does the confusion matrix represent in the context of classification
algorithms?
@

A) A matrix representing the confusion between the predicted and actual

classes
B) A matrix representing the confusion between different features in the
dataset
C) A matrix representing the confusion between the training and test datasets
D) A matrix representing the confusion between the training and validation
datasets
Accepted Answer: A) A matrix representing the confusion between the
predicted and actual classes

Page 1 MY SWAYAM
Introduction To Machine Learning IIT MADRAS

Question 20:
Which type of machine learning algorithm is inspired by the structure and
function of the human brain?
A) Decision Trees
B) Neural Networks
C) Support Vector Machines
D) K-Means Clustering

AM
Accepted Answer: B) Neural Networks

Question 21:
What is the purpose of the term "bias" in machine learning models?

AY
A) To introduce randomness into the model
B) To decrease the complexity of the model
C) To introduce a constant value to shift the predictions
SW
D) To remove outliers from the dataset

Accepted Answer: C) To introduce a constant value to shift the predictions

Question 22:
Y

Which algorithm is suitable for solving multi-class classification problems?

M

A) Logistic Regression
B) Decision Trees
C) Support Vector Machines
@

D) Random Forest

Accepted Answer: C) Support Vector Machines

Page 2 MY SWAYAM
Introduction To Machine Learning IIT MADRAS

Question 23:
What is the primary challenge of working with high-dimensional data in
machine learning?
A) It makes the data easier to analyze
B) It leads to faster model training
C) It leads to the curse of dimensionality, making algorithms prone to
overfitting
D) It improves the accuracy of machine learning models

AM
Accepted Answer: C) It leads to the curse of dimensionality, making
algorithms prone to overfitting

Question 24:

algorithms? AY
What is the role of the learning rate in gradient descent optimization

A) It defines the number of iterations for convergence

B) It controls the step size during each iteration to update the model
SW
parameters
C) It determines the size of the training dataset
D) It defines the number of features in the dataset

Accepted Answer: B) It controls the step size during each iteration to update
Y

the model parameters

M

Question 25:
What does PCA (Principal Component Analysis) do in machine learning?
@

A) It preprocesses textual data

B) It visualizes high-dimensional data in lower dimensions while preserving
its variance
C) It measures the accuracy of classification algorithms
D) It evaluates clustering algorithms

Accepted Answer: B) It visualizes high-dimensional data in lower dimensions

while preserving its variance

Page 3 MY SWAYAM
Introduction To Machine Learning IIT MADRAS

Question 26:
What does the term "bagging" refer to in ensemble learning?
A) A technique to prevent overfitting in decision trees
B) A technique to create multiple subsets of the dataset and train models
independently on each subset
C) A technique to boost the performance of neural networks
D) A technique to increase the learning rate of machine learning algorithms

AM
Accepted Answer: B) A technique to create multiple subsets of the dataset
and train models independently on each subset

Question 27:

AY
What is the primary goal of outlier detection in machine learning?
A) To increase the accuracy of the model
B) To identify and remove data points that do not follow the general trend of
SW
the dataset
C) To increase the complexity of the model
D) To identify and keep all data points for analysis

Accepted Answer: B) To identify and remove data points that do not follow the
Y

general trend of the dataset

M

Question 28:
What is the purpose of data normalization in machine learning?
@

A) To increase the range of the data

B) To decrease the range of the data
C) To add noise to the data
D) To remove outliers from the data

Accepted Answer: B) To decrease the range of the data

Page 4 MY SWAYAM
Introduction To Machine Learning IIT MADRAS

Question 29:
Which machine learning algorithm is used for outlier detection?
A) K-Means Clustering
B) Support Vector Machines
C) Isolation Forest
D) Naive Bayes

Accepted Answer: C) Isolation Forest

AM
Question 30:
What is the primary purpose of feature scaling in machine learning?

AY
A) To increase the range of features
B) To decrease the range of features
C) To add more features to the dataset
D) To remove features with low variance
SW

Accepted Answer: B) To decrease the range of features

Question 31:
Y

What is the difference between "accuracy" and "precision" in classification

metrics?
M

A) They are the same and can be used interchangeably

B) Accuracy measures the ratio of true positive predictions to the total
number of predicted positives, while precision measures the ratio of true
@

positive predictions to the total number of actual positives

C) Accuracy measures the ratio of true positive predictions to the total
number of actual positives, while precision measures the ratio of true positive
predictions to the total number of predicted positives
D) Accuracy measures the number of features, while precision measures the
number of data points

Accepted Answer: B) Accuracy measures the ratio of true positive predictions

to the total number of predicted positives, while precision measures the ratio

Page 5 MY SWAYAM
Introduction To Machine Learning IIT MADRAS

of true positive predictions to the total number of actual positives

Question 32:
What is the purpose of a confusion matrix in machine learning?
A) To create confusion among developers
B) To confuse machine learning algorithms
C) To evaluate the performance of a classification algorithm by summarizing
the actual and predicted classes
D) To visualize complex datasets

AM
Accepted Answer: C) To evaluate the performance of a classification
algorithm by summarizing the actual and predicted classes

AY Subscribe for
SW
PART 4

MY SWAYAM
Y

PDF OF ALL ABOVE QUESTIONS IS AVAILABLE IN THE

M

DESCRIPTION BOX OF THIS VIDEO.

FOR NEXT PART COMMENT BELOW AND SUBSCRIBE TO THE

@

“MY SWAYAM” YOUTUBE CHANNEL

HAPPY LEARNING
THANK YOU.

Page 6 MY SWAYAM