6/7/24, 4:08 PM ML Final Assessment-2

ML Final Assessment-2 Time :- 59:59 Min

1) In SVM, what is a hyperplane?

A) decision boundaries
B) data points
C) features
D) none of these

2) What is the primary goal of Data Science?

A) Making predictions based on historical data.
B) Extracting meaningful insights from data.
C) Developing complex algorithms
D) Designing databases

3) In Decision tree pruning methods include?

A) Prunning
B) Boosting
C) Bagging
D) All of the above

4) What is Machine learning?

A) The autonomous acquisition of knowledge through the use of computer pro
gram
B) The autonomous acquisition of knowledge through the use of manual progr
ams
C) The selective acquisition of knowledge through the use of computer progra
ms
D) The selective acquisition of knowledge through the use of manual programs

5) Decision Tree is a display of an algorithm?

A) True

https://wingz.itvedant.com/index.php/assignment-test-student/test-paper-topic-self?id=145882 1/10
6/7/24, 4:08 PM ML Final Assessment-2

B) False

6) Decision tree can be used for ______.

A) classification
B) regression
C) Both
D) None of these

7) Which of the following is a common error measure?

A) Median absolute deviation
B) Sensitivity
C) Specificity
D) All of the mentioned

8) __________________ uses Harmonic Mean in place of Arithmetic Mean as it

punishes the extreme values more.
A) S-measure
B) F-measure
C) T-measure
D) None of these

9) What do you mean by a hard margin?

A) The SVM allows very low error in classification
B) The SVM allows high amount of error in classification
C) The SVM allows no error in classification
D) none of these

10) Support Vector Machine(SVM) can be used for both classification or regression
challenges. Which kind of learning technique SVM uses?
A) unsupervised
B) supervised
C) reinforced
D) clustered

11) Generally, non-constant variance arises in presence of outliers or extreme

leverage values. Look like, these values get too much weight, thereby
disproportionately influences the models performance. The presence of non-
constant variance in the error terms results in _________?
A) heteroskedasticity
B) homoskedasticity
https://wingz.itvedant.com/index.php/assignment-test-student/test-paper-topic-self?id=145882 2/10
6/7/24, 4:08 PM ML Final Assessment-2

C) collinearity
D) diffusion

12) If a classification predictive model made 10 predictions and 5 of them were

correct and 2 of them were incorrect, then the classification accuracy of the model
based on just these predictions would be:

accuracy = correct predictions / total predictions * 100

accuracy = 5/ 10 * 100

accuracy = 50%

A) TRUE
B) FALSE

13) Which of the following are the Applications of Machine learning ?

A) Email Filtering
B) Sentiment Analysis
C) Face Recognition
D) All of the above

14) What is overfitting in the context of machine learning?

A) The model captures the underlying patterns in the data
B) The model performs well on the training data but poorly on unseen data
C) The model is too simple and fails to capture the complexity of the data
D) The model fits the noise in the training data.

15) Which of the following is/are true about boosting trees?

1.In boosting trees, individual weak learners are independent of each other

2.It is the method for improving the performance by aggregating the results of weak
learners

A) 2

https://wingz.itvedant.com/index.php/assignment-test-student/test-paper-topic-self?id=145882 3/10
6/7/24, 4:08 PM ML Final Assessment-2

B) 1
C) 1 & 2
D) 1^2

16) What is the primary goal of SVM in Machine Learning?

A) To maximize the margin between classes
B) To minimize the number of support vectors
C) To Visualize data relationships
D) To perform Unsupervised Learning

17) What does the assumption of Linearity in machine learning imply?

A) The relationship between features and target variable is not linear
B) The model should be a straight line
C) The residuals should be normally distributed
D) The data should be linearly separable

18) in the case of AdaBoost, very short decision trees were used that only had a
single split, called a ________________?
A) optimizer step
B) Decision stump
C) weak learner
D) bench step

19) Which graph is best suited for getting the below details?

a)distribution of the data Whether the data is normally distributed or if its skewed
(to the left or right)

b)To identify outliers Extremely low or high values that do not fall near any other
data points.

A) bar graph
B) histogram
C) scatter graph
D) pie chart

20) Function which performs the role of an activation function in machine learning
which is used to add non-linearity in a machine learning model. Basically, the
function determines which value to pass as output and what not to pass as output.

https://wingz.itvedant.com/index.php/assignment-test-student/test-paper-topic-self?id=145882 4/10
6/7/24, 4:08 PM ML Final Assessment-2

A) cost function
B) signoidal function
C) sigmoid function
D) none of these

21) Which of the following is correct use of cross validation?

A) Selecting variables to include in a model
B) Selecting parameters in prediction function
C) Comparing predictors
D) All of these

22) Unlike in AdaBoost, the incorrect result is given a higher weightage in gradient
boosting.
A) TRUE
B) FALSE

23) In linear regression, we try to ______ the least square errors of the model to
identify the line of best fit.
A) Change
B) Maximize
C) Minimize
D) None of the above

24) _________lies between Supervised and Unsupervised techniques.

A) association
B) clustering
C) semi supervised
D) None of these

25) A minmax scaler converts a distribution, such that it has range 0-1. The formula
used is_____?
A) Xsc=X−Xmin/(Xmax−Xmin.)
B) Xsc=X−Xmax−Xmin.
C) Xsc=X−Xmin−Xmin.
D) Xsc=X−Xmin.

26) Which regularization technique is used to penalize complex models?

A) L1 regularization
B) L2 regularization
C) Both A and B
https://wingz.itvedant.com/index.php/assignment-test-student/test-paper-topic-self?id=145882 5/10
6/7/24, 4:08 PM ML Final Assessment-2

D) None of these

27) Is XGBoosting Scalable?

A) YES
B) NO

28) For Lasso Regression, if the regularization parameter = 0, what does it mean?
A) The loss function is as same as the ordinary least square loss function
B) Can be used to select important features of a dataset
C) Shrinks the coefficients of less important features to exactly 0
D) All of the above

29) What’s the cost function of the logistic regression?

A) Sigmoid function
B) Logistic Function
C) both (A) and (B)
D) none of these

30) In Which approach, the breakdown is set to a binary classifier per each class?
A) One-to-Rest
B) One-to-Many
C) One-to-One approach
D) None of these

31) Ridge regression uses which norm?

A) L1
B) L2
C) Combination of L1 and L2
D) None

32) ____________loss is mostly used in SVM, this is used in the combination of

the activation function in the last layer. We use this loss to classify whether an email
is a spam or not.
A) Hinge loss
B) entropy loss
C) cross-entropy loss
D) MAE loss

33) Below lists some heuristics for best preparing your data for AdaBoost. Which of
the following is an approach?

https://wingz.itvedant.com/index.php/assignment-test-student/test-paper-topic-self?id=145882 6/10
6/7/24, 4:08 PM ML Final Assessment-2

A) Quality Data
B) Outlier
C) Noisy Data
D) All of these

34) Data Analysis is a process of?

A) inspecting data
B) cleaning data
C) transforming data
D) All of the above

35) Assuming that we have a dataset with little noise, a model is underfitting when:
A) Both the train & test errors are high.
B) Train error is low but test error is high.
C) Train error is high but test error is low.
D) Both train & test errors are low.

36) The term "Correlation" in EDA refers to ?

A) Measure of central tendency
B) Measure of spread of data
C) Measure of relationship between variables
D) Measure of data distribution

37) Which method in pandas provides a concise summary of a DataFrame,

including the number of non-null values in each column?
A) df.describe()
B) df.info()
C) df.head()
D) df.shape()

38) divide the total number of correctly classified positive examples by the total
number of predicted positive examples. Which is this formula?
A) Precision
B) Accuracy
C) Recall
D) None of these

39) Which of the following of the coefficients is added as the penalty term to the
loss function in Lasso regression?
A) Absolute value of magnitude
https://wingz.itvedant.com/index.php/assignment-test-student/test-paper-topic-self?id=145882 7/10
6/7/24, 4:08 PM ML Final Assessment-2

B) Squared magnitude
C) Number of non-zero entries
D) None of the above

40) Which of the following is a benefit of Gradient Boosting?

A) A benefit of the gradient boosting framework is that a new boosting algorith
m does not have to be derived for each loss function that may want to be used
B) A gradient descent procedure is used to minimize the loss when adding tree
s.
C) Both A and B
D) None of these

41) SVM is a supervised Machine Learning can be used for______?

A) Classification
B) Regression
C) Either a or b
D) None of these

42) Smart Speakers: Amazon Echo and Google Home are applications of machine
learning?
A) True
B) False

43) A weak classifier (decision stumps) is prepared on the training data

using____________?
A) random samples
B) average sample
C) weighted samples
D) None of these

44) Which one is true?

A) Ridge and Lasso regression are techniques to reduce the model complexity
and prevent over-fitting which may result from simple linear regression
B) Ridge regression shrinks the coefficients and it helps to reduce the model co
mplexity and multi-collinearity.
C) Lasso regression not only helps in reducing over-fitting but it can help us in f
eature selection
D) All of the above

45) What is the main purpose of Ridge and Lasso regularization in linear regression?
https://wingz.itvedant.com/index.php/assignment-test-student/test-paper-topic-self?id=145882 8/10
6/7/24, 4:08 PM ML Final Assessment-2

A) To increase the complexity of the model

B) To reduce the impact of outliers
C) To penalize large coefficients and prevent overfitting
D) To simplify the model by removing unnecessary features

46) The function to fill in NAN values with 10,if df is the dataframe object
A) fillna(10)
B) df.fill(10)
C) fillna(10)
D) df.fillna(10)

47) What is the primary purpose of a Loss Function in Linear Regression?

A) To maximize the accuracy of predictions
B) To minimize the difference between predicted and actual values
C) To calculate the mean of the target variable
D) To identify outliers in the data

48) Which of the following are real world applications of the SVM?
A) Text and Hypertext Categorization
B) Image Classification
C) Clustering of News Articles
D) All of the above

49) _______________is an implementation of gradient boosted decision trees

designed for speed and performance.
A) Ada Boost
B) Gradient Boost
C) XGBoost
D) All of these

50) Linear Regression is an example of?

A) Supervised Learning
B) Unsupervised Learning
C) Semi-Supervised Learning
D) Reinforcement Learning

https://wingz.itvedant.com/index.php/assignment-test-student/test-paper-topic-self?id=145882 9/10
6/7/24, 4:08 PM ML Final Assessment-2

SUBMIT

https://wingz.itvedant.com/index.php/assignment-test-student/test-paper-topic-self?id=145882 10/10