Machine learning MCQs

Machine learning MCQs

The following Machine learning MCQs have been compiled by our experts through research, in order to test your knowledge of the subject of Machine learning. We encourage you to answer these multiple-choice questions to assess your proficiency.
Please continue by scrolling down.

1:
In relation to machine learning, XGBoos it provides interfaces for which of the following options?
i) C++
ii) R
iii) Python
iv) Julia

A.  

All options 1). ii). iii). and iv)

B.  

Only options i). ii). and iii)


C.  

Only options i) and in

D.  

Only options i) and iii)

2:

What is the time complexity for breadth-f‌irst search algorithm? Given, d = depth.

A.  

bd 

B.  

O bd/2

C.  

O bd/3

D.  

O bd/4

3:

What will be the output of the following code in MATLAB?
z = [3 4 -2 6]:
Q = [6 8; 9 -4];
polyva|m(z, Q)

A.  

ans:
1802 3448
1754 158  


B.  

ans =
2802 2448
2754 -258 

C.  

ans =
2726 2536
2759 -254

D.  

ans =
2815 2496
2704 -222

4: Which of the following statements is/are correct about an Artif‌icial Neural Network (ANN)?

A.   The behaviour of an ANN is dependent on the weights that are specified for the units.

B.   The behaviour of an ANN is dependent on the transfer function that is specif‌ied for the units.

C.   Only (Dis correct.

D.   Only (ii)is correct.

E.   Both (i)and (ii)are correct.

F.   Neither (i)nor (ii)is correct

5: Which of the following statements is/ are correct about the Vowpal Wabbit?

A.   It does not allow the use of non-linear features.

B.   It can be used as a machine learning option in Microsoft Azure.

C.   Both options a and b

D.   Neither option 8 nor b

6:

In relation to machine learning, which of the following options is the DevOps tool for many real time monitoring dashboards of time series metrics?


A.  

Grafana 

B.  

Kibana

C.  

JupyterLab

D.  

Zepplein

7: Which of the following capabilities do Expert Systems possess?

A.   Substituting human decision makers.

B.   Predicting results.

C.   Refining their own knowledge

D.   Possessing human capabilities.

8: Which of the following is incorrect about machine learning?

A.   It does not require a human expert or programmer.

B.   It is cheap and f‌lexible.

C.   It does not require labeled data.

D.   It is used in pattern recognition.

9:

Which of the following methods attempts to fit the new predictor to the residual errors that has been made by the previous predictor?


A.  

AdaBoost 

B.  

Gradient Boosting

C.  

None of the above.

10:

Which of the following options is a Python library that allows the user to def‌ine, optimize, and evaluate the mathematical expressions, especially the ones with multi-dimensional arrays?


A.  

Torch 

B.  

Theano 

C.  

Shogun

D.  

Pattern

11: Simulation is appropriate under which of the following conditions?

A.   If cost exceeds savings.

B.   If a plan is visualized with animated simulation.

C.   If resource/time is not available.

D.   If system behavior is very complex.

12: Which of the following is a machine learning tool with application programming interface for C?

A.   Deeplearning4j

B.   LIBSVM

C.   Waff‌les

D.   GoLearn

13: Verification uses gray-box testing method.

A.   True

B.   False

14: In relation to machine learning classif‌ication, which of the following options is the correct way for defining recall?

A.   True Positive/ (True Positive + False Negative)

B.   True Positive/ (False Positive + False Negative)

C.   True Positive/ (True Positive + False positive)

D.   True Positive/ (True Negative + False Negative)

15: Which of the following types of data can be handled by the decision tree algorithm used for classification?

A.   Numerical data

B.   Categorical data

C.   Both numerical and categorical data

16:

in relation to machine learning classification, which of the following is an ensemble algorithm based on bagging?


A.  

Random Forest 

B.  


Naive Bayes


C.  

Decision Trees

D.  

SVM

17:

Which of the following elements of a typical reinforcement learning algorithm def‌ines the goals of a learner as a numerical value?


A.  

A model of the environment.

B.  

A value function.   

C.  

A reward function.   

D.  

A policy.

18: Which of the following simulation tools rely on a transaction-flow approach toward modeling systems?

A.   Hybrid simulator

B.   Continuous simulator

C.   Discrete event simulator

D.   Agent-based simulator

19: What is the accuracy of the Logistic Regression classif‌ication algorithm?

A.   80.11%

B.   82.53%

C.   84.60%

D.   86.27%

20:

Which of the following machine learning techniques are implemented by Apache Mahout?


(i)Classification


(ii)Clustering


(iii)Recommendation


A.  

Only (nand (ii)

B.  

Only (ii)and (iii) 

C.  

Only (f‌iand (iii)

D.  

All (i), (ii)and Gil)

21: Which of the following is/ are the large margin classifier(s)?

A.   SVM

B.   Boosting

C.   Both SVM and Boosting

22:

What will be the output of the following code in MATLAB?

M=(864;372;193]

N = circshift(M,[-1 2])


A.  

M=

8 6 4

3 7 2

1 9 3

N=

2 3 7

3 1 9

4 8 6 


B.  

M:

8 6 4

3 7 2

1 9 3

N=

7 2 3

9 3 1

6 4 8 

C.  

M:
864
372
193
N:
732
913
684

D.  

M:
864
372
193
N:
139
327
846

23:

What will be the output of the following code in MATLAB?

Z=[428;735;612]

sort(Z, 1)


A.  

 Z:

4 2 8

7 3 5

6 1 2

ans=

2 4 8

3 5 7

1 2 6

B.  

 z:

4 2 8

7 3 5

6 1 2

ans=

4 2 8

6 1 2

7 3 5  


C.  

z:

428

735

612

ans=

412

625

738

D.  

Z  =

428

735

612

ans=

126

248

357


24: With respect to Data Mining Query Language, which of the following is the syntax for characterization?

A.   mine characteristics [ as [pattern_name)]

B.   {matching {metapattern}}

C.   mine characteristics [as pattern_name]

D.   analyze classifying_attribute_or_dimension

E.   O mine characteristics [as pattern_name] analyze [measure(s)}

F.   mine Characteristics [as pattern_name] analyze prediction_attribute_or_dimension

G.   {set [attribute_or_dimension_i= value_i}]

25:

Which of the following data warehousing approaches is/are used to integrate heterogeneous databases?

(i)Ouery-driven approach

(ii)Update-driven approach


A.  

Only (i)   

B.  

Only (ii) 

C.  

Both (i)and (ii)

D.  

Neither (i)nor (ii)

26:

Which of the following SimPy function calls is/are used in marking a thread as runnable when it is first created?


A.  

activateO 

B.  

simulateo

C.  

reactivateo

D.  

Both a and b

27: Which of the following is NOT an advantage of simulation modelling and analysis?

A.   Multiple performance metrics to analyze system configurations.

B.   Identification of bottlenecks in the flow of information.

C.   Test hypotheses of the system for feasibility.

D.   Excellent schedule and budget planning.

E.   Decision tree algorithm is a type of:

F.   Decision tree algorithm is a type of:

G.   Supervised learning.

H.   unsupervised learning.

I.   reinforcement learning.

28: Which of the following statements is/are correct about unsupervised machine learning?

A.   Its algorithms learn to predict the output from input data.

B.   Its algorithms learn to inherent the structure from input data.

C.   It is used against data that has historical labels.

D.   Both b and c are correct.

29: Logistic regression can be used with which of the following types of dependent variables?

A.   Integer

B.   Float

C.   Binary (Yes/No response)

D.   All of the above.

30: Which ofthe following boosting algorithms implements regularization that helps in reducing overfitting?

A.   XGBoost

B.   AdaBoost

C.   Gradient Boost

D.   Neither of the above.

31: What is the accuracy of Naive Bayes algorithm used for classif‌ication?

A.   78.93%

B.   83.25%

C.   80.11%

D.   86.72%

32:

Which of the following SimPy operations is used for indicating the passage of a certain amount of time within a thread?


A.  

yield request

B.  

yield release 

C.  

yield hold    

D.  

yield passivate

33:

Which of the following is a Python tool that extends the functionality of NumPy and SciPy packages and provides functions for performing classif‌ication. regression, clustering and dimensionality reduction,

model selection. and preprocessing?


A.  

RapidMiner  

B.  

Weka3 

C.  

Scikit-Learn    

D.  

Shogun

34: Which of the following statements is/ are correct about the C45 classification technique?

A.   It can deal with both continuous features as well as discrete features.

B.   It can be used for small as well as large datasets.

C.   It has low processing time.

D.   All of the above.

35: Which of the following algorithms does not scale up well as compared to the other for very large datasets?

A.   LibLinear

B.   Vowpal Wabbit

C.   LibSVM

D.   All of the above scale very well for large datasets.

36: What is the function of ridge regression?

A.   It uses L1 regularization.

B.   It is used when data suffers from multi-collinearity.

C.   It shrinks coefficients to zero.

D.   It uses absolute values in the penalty function.

37: Simulation languages offer less flexibility than application-oriented simulators.

A.   True

B.   False

38:

Which of the following is NOT a characteristic of the NlCeSim simulator, which uses machine learning techniques?


A.  

Flexible 

B.  

Static 

C.  

Open-source

D.  

Dynamic

39: Which of the following statements is correct about validation?

A.   It uses black-box and white-box testing.

B.   It does not execute a code.

C.   It involves human-based checking of files and documents.

D.   It is a static mechanism.

40: In relation to artificial neural network, which of the following options is designed for working with sequence prediction problems?

A.   Multilayer Perceptrons (MLPs)

B.   Convolutional Neural Networks (CNNs)

C.   Recurrent Neural Networks (RNNs)

D.   None of the above.

41:

In relation to artificial neural network, the Recurrent Neural Networks (RNN), should NOT be used for which of the following types of data?


A.  

Speech data  

B.  

Generative Model 

C.  

Tabular data  

D.  

Text data

42: Which of the following is NOT applicable to speech recognition application in machine learning?

A.   It is used in hands-free computing.

B.   It aims to understand and comprehend the spoken word.

C.   It is used in menu navigation.

D.   It is speaker-dependent.

43: ln machine learning, logistic regression:

A.   requires a linear relationship between independent and dependent variables.

B.   supports multi-collinearity.

C.   requires small sample sizes.

D.   is used for classif‌ication problems.

44:

Which of the following techniques can be used for solving the attribute conditional density estimation problem in the Bayesian Network classif‌ication method?


A.  

Decision-tree structured conditional probability   

B.  

Greedy learning algorithm  

C.  

Prototype selection

D.  


Gaussian kernel function  


45:

Choose the answer to the following question from the options given in the image.

In machine learning, which of the given functions in the image is used to measure the size of vectors?

A.  


B.  

B

C.  

C

D.  

46:

In the SVM classif‌ication approach, which of the following options is used for solving the issue of the low-sparse SVM classifier?


A.  

Risk area SVM 

B.  

Fuzzy SVMs 

C.  

Cluster Support Vector Machine   

D.  

Prototype selection

47: Which of the following defines the F-score measure to assess the quality of text retrieval?

A.   F-score = recall x precision / (recall - precision)/ 2

B.   F-score = |[Relevant} f‌l {Retrieved}l / l[Retrieved]l

C.   F-score = |{Relevant} fl [Retrieved}l / l[Relevant]|

D.   F-score = recall x precision / (recall + precision)/ 2

48:

Which of the following tools is used for processing, analyzing and visualizing the large data sets and can provide native support for Apache Spark distributed computing?


A.  

Zeppelin 

B.  

Jupyter

C.  

Kibana

D.  

Tableau

49: Which of the following is NOT a multi-class classif‌ier?

A.   Classif‌ication of spam and non-spam emails

B.   Classif‌ication of crop types

C.   Classif‌ication of mood

D.   All of the above are multi-class identif‌iers.

50: A multi-layer perceptron (MLP)is a:

A.   finite cyclic graph.

B.   inf‌inite cyclic graph

C.   inf‌inite acyclic graph

D.   finite acyclic graph.

51: Which of the following is NOT an example of bounded probability distribution?

A.   Uniform distribution

B.   Triangular distribution

C.   Logistic distribution

D.   Binomial distribution

52:

Which of the following methods is used to generate non-uniform random variates and uses multiple uniform [0,1] variables?


A.  

Convolution 

B.  

Composition 

C.  

Inverse Transform

D.  

Acceptance-Rejection

53:

Which of the following machine learning tools supports vector machines, dimensionality reduction, and online learning, etc.?


A.  

Colab 

B.  

Shogun 

C.  

Accors.Net

D.  

Weka

54: Simple linear regression is characterized by how many independent variables?

A.   One

B.   Two

C.   Three

D.   Four

55: Which of the following machine learning tools provides API for the neural networks?

A.   Keras.io

B.   Accors.net

C.   Rapid Miner

D.   Shogun

56: Which of the following machine learning frameworks works at the higher level of abstraction?

A.   Tensorf‌low

B.   PyTorch

C.   Theano

D.   Keras

57: Which of the following types Of data analysis models is/are used to conclude continuous valued functions?

A.   Prediction

B.   Classif‌ication

C.   Both a and b

D.   Neither a nor b

58:

In relation to K-Nearest Neighbors (K-NN) algorithm, what effect does the small k value (neighborhood size) have on bias and variance?


A.  

High Bias, High Variance

B.  

Low Bias, Low Variance  

C.  

High Bias, Low Variance

D.  

Low Bias, High Variance

59:

In the knowledge discovery process, which of the following steps is involved in retrieving data, relevant to the analysis task, from the database?


A.  

Data selection

B.  

Data mining

C.  

Data transformation

D.  

Knowledge presentation

60: Which of the following is an advantage in logistic regression?

A.   Natural probabilistic view of class predictions.

B.   Linear decision boundary.

C.   Independent observations requirement.

D.   Overfitting the Model.

61: In relation to machine learning framework, which of the following languages is used by Veles for performing automation and coordination between the nodes?

A.   C++

B.   Java

C.   Python

D.   R

62: Which of the following is incorrect about linear regression?

A.   Independent variables in linear regression can be continuous or discrete.

B.   A dependent variable in linear regression is discrete.

C.   Linear regression is sensitive to outliers.

D.   It estimates the real values that are based on continuous variables.

63: Which of following is incorrect about modelica?

A.   It is a language

B.   It is a tool.

C.   It is declarative.

D.   It is object-oriented.

64: Which of the following statements is incorrect about Recurrent Neural Networks (RNNs)?

A.   They are non-deterministic.

B.   They can settle to point attractors.

C.   They can oscillate.

D.   They have at least one feed-back connection.

65: In relation to machine learning. the XGBoost works for which of the following options?

A.   Computer vision

B.   Tabular data

C.   NLP

D.   All of the above.

66: In Multilayer Perceptrons, the predictions are made in which of the following layers?

A.   Input layer

B.   First hidden layer

C.   Last hidden layer

D.   Output layer

67:

Any layer that the user wants to use.

In agent-based simulation, which of the following Inferential Theory of Learning (ITL)operations is used to modify knowledge by narrowing the reference set of a description?


A.  

Concretion 

B.  

Generalization 

C.  

Specialization 

D.  

Abstraction

68: Which of the following is incorrect about the feedback artif‌icial neural network?

A.   They are used in content addressable memories.

B.   Feedback loops are allowed.

C.   Flow of information is bi-directional.

D.   Feedback networks are static.

69: Which of the following is incorrect about the feedforward artificial neural network topology?

A.   Feedback loops are not allowed.

B.   Flow of information is unidirectional.

C.   They do not have fixed inputs and outputs.

D.   They are used in pattern generation.

70: The machine learning framework RapidMiner is written in which of the following programming languages?

A.   C++

B.   Java

C.   Python

D.   C

71:

Which of the following machine learning tools works with large data volume and supports text mining and image mining through plugins?


A.  

KNIME 

B.  

Weka

C.  

Colab

D.  

TensorFIow

72:

Which of the following statements is true?

Statement 1: Reinforcement learning is an off-line technique.

Statement 2: The reinforcement learning technique is used in elevator scheduling.

A.   Statement1 is true.

B.   Statement 2 is true.

C.   Both statements 1 and 2 are true.

D.   Both statements 1 and 2 are false.

73:

What will be the output of the following code, when executed in MATLAB?

Z=[8471;7196;6429]

Z(3:3,1:3)


A.  

2:

8 4 7 1

7 1 9 6

6 4 2 9

ans=

8 4 7 

B.  

 Z:

8 4 7 1

7 1 9 6

6 4 2 9

ans:
7 1 9   


C.  

Z:

8 4 7 1

7 1 9 6

6 4 2 9

ans:
6 4 2  


D.  

OZ:
8 4 71
7 1 9 6
6 4 2 9
ans=
4 7 1


74:

Which of the following are the applications of data mining?


(i)Science Exploration


(ii)Fraud Detection


(iii)Customer Retention


A.  

Only (i)and (ii)

B.  

Only (ii)and (iii)  

C.  

Only (i)and (iii)

D.  

All (i), (ii)and (iii)

75: An artificial neuron device consists of how many inputs and outputs?

A.   One input and many outputs.

B.   Many inputs and one output

C.   One input and one output.

D.   Many inputs and many outputs.

76:
In machine learning, random forests method is an example of which of the following feature selection methods?

A.  

Wrapper method  

B.  

Embedded method

C.  

Filter method

77:

A target in machine learning is known as alan:

A.  

label

B.  

dependent variable.

C.  

object

D.  

feature

78:

In relation to classif‌ication in machine learning, which Of the following Options is the correct way for def‌ining the F1-Score?


A.  

F1-Score: (Precision x Recall) / ((2 x Precision) + Recall)     

B.  

F1-Score: (2 x Precision x Recall) / (Precision + Recall)     

C.  

F1-Score: (Precision x Recall) / 2(Precision + Recall)

D.  

F1-Score: (Precision x Recall) / (Precision 1‘ (2 x Recal|))

79:

What will be the output of the following code, when executed in MATLAB?

X = 45;

Y = 21;

Z = bitor(X, Y)

Z = bitxor(X, Y)

Z = bitshift(X,-3)

Z = bitshift(X,4)


A.  

2:63

2:59

2:6

2:71 


B.  

 Z=62

2: 59

2:8

2:73 


C.  

2:69

2: 54

2:7

2:720


D.  

2:61

2:56

2:5

2:72


80:

In relation to machine learning classif‌ication, which of following options refers to the graphical model for probability associations between a set of variables?


A.  

Bayesian Network

B.  

K—Nearest Neighbors


C.  

lbs

D.  

SVM

81:

What will be the output of the following code, when executed in MATLAB?

z = [6 9 4 3 5]:

polyval(z,3)


A.  

ans= 779

B.  

ans= 773

C.  

ans = 821

D.  

ans= 782

82:
Which Of the following statistical data mining techniques is used to predict a categorical response variable?

A.  

Mixed-effect model 

B.  

Factor analysis 

C.  

Regression

D.  

Generalized linear model

83:

Consider the following code to be executed in MATLAB.

m = roots([?., 8])

What will be the output?


A.  

m = 4  


B.  

 m = -4  

C.  

m =16

D.  

 m = -16

84: The EIasticNet regression technique:

A.   cannot suffer with double shrinkage.

B.   discourages group effect in case of highly correlated variables.

C.   is a hybrid of lasso and linear regression techniques.

D.   does not have any limitations on the number of selected variables.

85:

Find the output of the following code, when executed in MATLAB.

A = [8 4 7 5 3]

isinteger(A)

isf‌loat(A)

isvector(A)

isscalar(A)


A.  

A =

8 4 7 5 3
ans = 0

ans = 1

ans =1

ans =      

B.  

A =

8 4 7 5 3
ans =1

ans = 0

ans = 0

ans =1

C.  

A:
8 4 7 5 3
ans=0
ans=1
ans=0
ans=1

D.  

A:
8 4 7 5 3
ans=1
ans=1
ans=0
ans=0

86:

Which of the following statements is/are correct?

Statement 1: In stochastic simulation environments, machine learning is performed by combining multiple transmutations.

Statement 2: Stochastic simulation is used for modelling a system whose operation can be directly captured by deterministic rules.

A.  

Statement 1 is correct.  

B.  

Statement 2 is correct.

C.  

Both statements 1 and 2 are correct.

D.  

Both statements 1 and 2 are incorrect.

87: In data mining, which of the following is NOT a data reduction technique?

A.   Clustering

B.   Sampling

C.   Histograms

D.   Huffman

88: With respect to multilayer neural network, what does the neurons in the hidden layer correspond to?

A.   Properties

B.   Descriptors

C.   Non-linear latent variables

D.   Both a and c

89: Which of the following boosting algorithms uses the level-wise tree growth?

A.   XGBoost

B.   Gradient Boosting

C.   AdaBoost

D.   Light GB

90:

Which of the following algorithms has the given applications?

1. Scene classification

2. Induction motors fault diagnosis

3. Analog circuit fault diagnosis

4. Corporate financial distress prediction

A.  

04.5

B.  

SVM

C.  

ID3

D.  

Bayesian Network

91: Can decision trees over f‌it data?

A.   Yes

B.   No

92: Which of the following simulation model types contain(s)probability?

A.   Deterministic simulations

B.   Stochastic simulations

C.   Both a and b

D.   Neither a nor 0

93: What is the correct F1 score of the Random Forest algorithm of classif‌ication?

A.   0.5924

B.   0.5171

C.   0.6518

D.   0.6275

94: What is the FI-Score of Random Forest algorithm used for classification?

A.   0.6737

B.   0.5180

C.   0.5224

D.   0.6275

95: Which of the following interfaces is/are supported by Scikit—Learn tool?

A.   GPU

B.   API

C.   Both GPU and API

96: Which of the following formulae is used for correctly calculating the accuracy of the classification algorithms?

A.   Accuracy: 2(True Positive + True Negative) I Total Population

B.   Accuracy: (True Positive + True Negative) / Total Population

C.   Accuracy: (True Positive + True Negative) / 2(Total Population)

D.   Accuracy: (True Positive x True Negative) / Total Population

97: In data mining, which of the following clustering methods reflects spatial distribution of the data points?

A.   Grid-based method

B.   Model-based method

C.   Partitioning method

D.   Hierarchical method

98: In relation to SVM classification algorithm. which of the following options is used for the binary classification?

A.   Polynomial kernel

B.   Radial Basis Function kernel

C.   Sigmoid kernel

D.   None of the above.

99: Which of the following classif‌ication algorithms should be used for the given scenario? The data is labeled and the number of samples is greater than 100k.

A.   SGD Classif‌ier

B.   Kernel Approximation

C.   Linear SVC

D.   None of the above.

100: Which of the following supervised learning algorithms is/are implemented by Apache Mahout?

A.   Naive Bayes classifiers

B.   Support Vector Machines (SVMs)

C.   Neural Networks

D.   Both a and c