1. Which of the following uses Machine Learning to predict traffic?

A. Virtual Personal Assistant

B. Social Media Services
C. Traffic Prediction
D. Online Fraud Detection
Answer: C

2. Which of the following uses Machine Learning to personalize our news feed on Facebook?
A. Product Recommendation
B. Online Transportation
C. Email Spam Filtering
D. Social Media Services
Answer: D

3. Which of the following uses Machine Learning to compute the estimated cost and distance
of travel in apps like Uber?
A. Virtual Personal Assistant
B. Traffic Prediction
C. Online Transportation
D. Online Fraud Detection
Answer: C

4. Which of the following uses Machine Learning to filter email spam?

A. Product Recommendation
B. Email Spam Filtering
C. Social Media Services
D. Online Fraud Detection
Answer: B

5. Which of the following is not a commonly used Machine Learning programming

language?
A. Python
B. C++
C. Ruby
D. Scala
Answer: C
6. What is structured data?
a) Data with unknown structure
b) Data that can be stored, accessed and processed in fixed format
c) Data that contains a combination of simple text files, images, videos etc.
d) Data represented in an XML file.

Answer: b) Data that can be stored, accessed and processed in fixed format.
7. Which of the following is an example of structured data?
a) The output returned by a Google search
b) A heterogeneous data source containing a combination of simple text files, images, videos
etc.
c) An 'Employee' table in a database
d) Personal data stored in an XML file.
Answer: c) An 'Employee' table in a database.
8. What is unstructured data?
a) Data with unknown structure
b) Data that can be stored, accessed and processed in fixed format
c) Data that contains a combination of simple text files, images, videos etc.
d) Data represented in an XML file.

Answer: a) Data with unknown structure.

9. Which of the following is an example of unstructured data?

a) The output returned by a Google search
b) A heterogeneous data source containing a combination of simple text files, images, videos
etc.
c) An 'Employee' table in a database
d) Personal data stored in an XML file.

Answer: a) The output returned by a Google search.

10. What is semi-structured data?

a) Data with unknown structure
b) Data that can be stored, accessed and processed in fixed format
c) Data that contains a combination of structured and unstructured data
d) Data represented in an XML file.

Answer: c) Data that contains a combination of structured and unstructured data.

11. What is the meaning of the term "Volume" in the context of Big Data?
a) How fast data is produced
b) Different forms of data
c) How much data is generated
d) The quality or accuracy of captured data

Answer: c) How much data is generated.

12. What is the meaning of the term "Velocity" in the context of Big Data?
a) How fast data is produced
b) Different forms of data
c) How much data is generated
d) The quality or accuracy of captured data
Answer: a) How fast data is produced.

13. What is the meaning of the term "Variety" in the context of Big Data?
a) How fast data is produced
b) Different forms of data
c) How much data is generated
d) The quality or accuracy of captured data
Answer: b) Different forms of data.

14. What is the meaning of the term "Veracity" in the context of Big Data?
a) How fast data is produced
b) Different forms of data
c) How much data is generated
d) The quality or accuracy of captured data

Answer: d) The quality or accuracy of captured data.

15. What is structured data?
a) Data with unknown structure
b) Data that can be stored, accessed and processed in fixed format
c) Data that contains a combination of simple text files, images, videos etc.
d) Data represented in an XML file.
Answer: b) Data that can be stored, accessed and processed in fixed format.

16. Which of the following is an example of structured data?

a) The output returned by a Google search
b) A heterogeneous data source containing a combination of simple text files, images, videos
etc.
c) An 'Employee' table in a database
d) Personal data stored in an XML file.
Answer: c) An 'Employee' table in a database.

17. What is unstructured data?

a) Data with unknown structure
b) Data that can be stored, accessed and processed in fixed format
c) Data that contains a combination of simple text files, images, videos etc.
d) Data represented in an XML file.
Answer: a) Data with unknown structure.

18. Which of the following is an example of unstructured data?

a) The output returned by a Google search
b) A heterogeneous data source containing a combination of simple text files, images, videos
etc.
c) An 'Employee' table in a database
d) Personal data stored in an XML file.
Answer: a) The output returned by a Google search.

19. What is semi-structured data?

a) Data with unknown structure
b) Data that can be stored, accessed and processed in fixed format
c) Data that contains a combination of simple text files, images, videos etc.
d) Data represented in an XML file.
Answer: c) Data that contains a combination of structured and unstructured data.
20. What is Big Data analytics?
A) The process of collecting, organizing and analyzing small sets of data
B) The process of collecting, organizing and analyzing large sets of data to discover patterns
and other useful information
C) The process of collecting, organizing and analyzing data to create a database
D) The process of collecting, organizing and analyzing data to predict future trends
Answer: B) The process of collecting, organizing and analyzing large sets of data to discover
patterns and other useful information

21. What is the benefit of Big Data analytics?

A) To collect data from different sources
B) To identify the data that is most important to the business and future business decisions
C) To process small volumes of data
D) To store data in different places and systems
Answer: B) To identify the data that is most important to the business and future business
decisions

22. What kind of software tools and applications are used for Big Data analytics?
A) Traditional database and software methods
B) Applications for structured data only
C) Specialized software tools and applications for predictive analytics, data mining, text
mining, forecasting and data optimization
D) Applications for unstructured data only
Answer: C) Specialized software tools and applications for predictive analytics, data mining,
text mining, forecasting and data optimization
23. What is the biggest challenge of Big Data analysis for most organizations?
A) Identifying the data that is most important to the business
B) Combining different types of data to find patterns and other useful business information
C) Collecting data from different sources
D) Breaking down data silos to access all data an organization stores in different places and
often in different systems
Answer: D) Breaking down data silos to access all data an organization stores in different
places and often in different systems
24. Why is it difficult to process Big Data using traditional database and software methods?
A) Because the data is too small
B) Because the data is too simple
C) Because the data is too large
D) Because the data is too structured
Answer: C) Because the data is too large
25. Which of the following is the most effective method to predict future trends and patterns?
A) Ecommerce
B) Marketing
C) Analytics
D) Research
Answer: C) Analytics
26. Which of the following companies has helped millions of people to buy products from
any corner of the world?
A) Amazon
B) Google
C) Facebook
D) Walmart
Answer: A) Amazon
28. What is the main purpose of marketing?
A) To improve and facilitate online transactions
B) To predict future trends and patterns
C) To attract potential clients through various techniques
D) To understand consumer behavior better
Answer: C) To attract potential clients through various techniques
29. What is the main benefit of Customer Relationship Management?
A) To keep track of a customer's buying history and preferences
B) To understand consumer behavior better
C) To create a better relationship with the customers
D) To predict future trends and patterns
Answer: C) To create a better relationship with the customers
30. Which of the following is NOT a characteristic of Big Data?
A) Volume
B) Velocity
C) Variety
D) Validity
Answer: D) Validity (The correct characteristic is Veracity)
31. What is big data?
a) Data collected from a single source
b) Data collected from multiple sources and processed for various purposes
c) Data generated by desktop computers
d) Data generated by social media websites
Answer: b) Data collected from multiple sources and processed for various purposes
32. Who is a data scientist?
a) A person who collects data
b) A person who processes data
c) A person who analyzes and interprets complex data
d) A person who stores data
Answer: c) A person who analyzes and interprets complex data
33. What is the main source of data generation?
a) Desktop computers
b) Laptops
c) Mobile phones and tablets
d) Smartwatches
Answer: c) Mobile phones and tablets
34. What are the top 5 popular big data applications?
a) Ecommerce, Analytics, Marketing, Research, and Finance
b) Ecommerce, Analytics, Marketing, Research, and Customer Relationship Management
c) Ecommerce, Analytics, Finance, Research, and Customer Relationship Management
d) Analytics, Marketing, Research, Finance, and Customer Relationship Management

Answer: b) Ecommerce, Analytics, Marketing, Research, and Customer Relationship

Management
35. What is the "big data" initiative launched by the Obama administration?
a) Initiative to collect data from a single source
b) Initiative to process data for various purposes
c) Initiative to store data in a database
d) Initiative to advance research and development in big data
Answer: d) Initiative to advance research and development in big data
36. What is big data according to Gartner?
a) Low-volume, low-velocity, and/or low-variety information assets
b) High-volume, low-velocity, and/or high-variety information assets
c) Low-volume, high-velocity, and/or high-variety information assets
d) High-volume, high-velocity, and/or high-variety information assets
Answer: d) High-volume, high-velocity, and/or high-variety information assets
37. How can a small data be considered as big data?
a) By increasing its volume
b) By increasing its velocity
c) By applying complicated analysis
d) By increasing its variety
Answer: c) By applying complicated analysis
38. What impact has big data made in the world of IT?
a) No impact
b) Minor impact
c) Major impact
d) Negative impact
Answer: c) Major impact
39. What are the top 5 popular big data applications?
a) Ecommerce, Analytics, Marketing, Research, and Finance
b) Ecommerce, Analytics, Marketing, Research, and Customer Relationship Management
c) Ecommerce, Analytics, Finance, Research, and Customer Relationship Management
d) Analytics, Marketing, Research, Finance, and Customer Relationship Management
Answer: b) Ecommerce, Analytics, Marketing, Research, and Customer Relationship
Management
40. What is the main purpose of processing big data?
a) To store large amounts of data
b) To enable enhanced decision making, insight discovery and process optimization
c) To reduce the amount of data generated
d) To only process low-volume, low-velocity, and/or low-variety information assets
Answer: b) To enable enhanced decision making, insight discovery and process optimization
41. What is big data?
a) Data collected from a single source
b) Data collected from multiple sources and processed for various purposes
c) Data generated by desktop computers
d) Data generated by social media websites

Answer: b) Data collected from multiple sources and processed for various purposes
42. What has led to the growth of big data?
a) Decrease in the use of internet
b) Decrease in the use of smartphones
c) Advancement in technologies
d) Decrease in the use of cloud storage
Answer: c) Advancement in technologies
43. What is the main source of data generation?
a) Desktop computers
b) Laptops
c) Mobile phones and tablets
d) Smartwatches
Answer: c) Mobile phones and tablets
44. What has increased the amount of data generation?
a) Decrease in the use of social media websites
b) Decrease in the use of online shopping websites
c) Increase in the use of social media websites and online shopping websites
d) Decrease in the use of mobile devices
Answer: c) Increase in the use of social media websites and online shopping websites
45. What is big data an amalgamation of?
a) Information gathered from a single source
b) Information gathered from multiple sources
c) Information generated by desktop computers
d) Information generated by social media websites
Answer: b) Information gathered from multiple sources
46. What is machine learning?
a) An application of artificial intelligence that involves algorithms and data that automatically
analyse and make decisions without human intervention.
b) A process of manually programming a computer to perform tasks.
c) A way to navigate on Mars without human expertise.
Answer: a
47. How is artificial intelligence generated in machine learning?
a) By human developers giving codes that instruct the system how to react to situations.
b) By the computer performing tasks based on previous experiences.
c) By gathering data from human expertise.
Answer: b
48. What is the difference between normal computer software and machine learning?
a) Normal computer software is trained by a large number of data.
b) Machine learning is trained by a large number of data.
c) Normal computer software is programmed to react to situations.
Answer: c
49. When is learning used in machine learning?
a) When there is a need to calculate payroll.
b) When humans are unable to explain their expertise.
c) When the solution needs to be adapted to particular cases.
Answer: b
50. What is the role of statistics in machine learning?
a) To efficiently represent and evaluate the model for inference.
b) To solve the optimization problem.
c) To infer from a sample.
Answer: c
51. What is the role of computer science in machine learning?
a) To efficiently represent and evaluate the model for inference.
b) To solve the optimization problem.
c) To infer from a sample.
Answer: a
52. What is the purpose of building a model in machine learning?
a) To gather data from human expertise.
b) To optimize a performance criterion using example data or past experience.
c) To program a computer to react to situations.
Answer: b
53. Which of the following is not an advantage of Machine Learning?
A. Fast
B. Accurate
C. Requires human intervention
D. Efficient
Answer: C
54.What is one of the disadvantages of Machine Learning?
A. Fast processing time
B. Easy to rectify errors
C. Difficult data acquisition
D. Requires less time and space for result interpretation
Answer: C
55. What is Artificial Intelligence?
a) The study of algorithms that improve their performance at some task with experience.
b) A concept of creating intelligent machines that simulate human behaviour.
c) The process of labelling input data and expected output data.

Answer: b

56. What is Machine Learning?

a) A subset of Artificial Intelligence that allows machines to learn from data without being
programmed.
b) A technique where the program is given labelled input data and expected output data.
c) The process of improving performance at some task with experience.

Answer: a

57. What is Supervised Learning?

a) A technique where the program is given labelled input data and expected output data.
b) A technique where the program is given unlabelled input data.
c) A technique where the program is given rewards from a sequence of actions.

Answer: a

58. What is Unsupervised Learning?

a) A technique where the program is given labelled input data and expected output data.
b) A technique where the program is given unlabelled input data.
c) A technique where the program is given rewards from a sequence of actions.

Answer: b

59. What is Semi-Supervised Learning?

a) A technique where the program is given labelled input data and expected output data.
b) A technique where the program is given unlabelled input data.
c) A technique where the program is given a few labelled input data.

Answer: c

60. What is Reinforcement Learning?

a) A technique where the program is given labelled input data and expected output data.
b) A technique where the program is given unlabelled input data.
c) A technique where the program is given rewards from a sequence of actions.

Answer: c
61. What is Regression in Machine Learning?
a) A technique where Y is categorical (qualitative).
b) A technique where Y is continuous (quantitative).
c) A technique where the program is given rewards from a sequence of actions.

Answer: b

62. What is Classification in Machine Learning?

a) A technique where Y is categorical (qualitative).
b) A technique where Y is continuous (quantitative).
c) A technique where the program is given rewards from a sequence of actions.

Answer: a

63. What is Speech Recognition?

a) A technique where Y is categorical (qualitative).
b) A technique where Y is continuous (quantitative).
c) The process of identifying words spoken according to speech signals.

Answer: c

64. What is the purpose of Automatic voice recognition systems?

a) To predict the value of the stock market in 6 months.
b) To identify words spoken according to speech signals.
c) To classify input data into categories.

Answer: b
65. What is Clustering in Machine Learning?
a) A technique where Y is categorical (qualitative).
b) A technique where Y is continuous (quantitative).
c) Grouping similar instances without any pre-existing labels.

Answer: c

66. What is the main difference between Supervised and Unsupervised Learning?
a) Supervised Learning requires labelled input data, while Unsupervised Learning does not.
b) Unsupervised Learning requires labelled input data, while Supervised Learning does not.
c) Both require labelled input data.

Answer: a

67. What is Deep Learning?

a) A technique where the program is given labelled input data and expected output data.
b) A technique where the program is given unlabelled input data.
c) A type of Machine Learning that uses neural networks with multiple layers.

Answer: c

68. What is the Credit Assignment Problem in Reinforcement Learning?

a) The problem of assigning credit to actions that led to a delayed reward.
b) The problem of assigning credit to labelled input data.
c) The problem of assigning credit to unlabelled input data.

Answer: a

69. What is Overfitting in Machine Learning?

a) When the model performs well on the training data but poorly on the testing data.
b) When the model performs well on the testing data but poorly on the training data.
c) When the model performs equally well on both the training and testing data.

Answer: a
60. What is Big Data analytics?
a) The process of collecting, organizing, and analyzing small sets of data
b) The process of collecting, organizing, and analyzing large sets of data
c) The process of collecting, organizing, and analyzing medium-sized sets of data
Answer: b) The process of collecting, organizing, and analyzing large sets of data.

70. What can Big Data analytics help organizations with?

a) Better understanding of less important data
b) Identifying the data that is least important to the business
c) Better understanding of important data and identifying data that is most important to the
business and future business decisions
Answer: c) Better understanding of important data and identifying data that is most important
to the business and future business decisions.

71. What specialized software tools and applications are typically used for Big Data
analytics?
a) Spreadsheet software
b) Word processing software
c) Specialized software tools and applications for predictive analytics, data mining, text
mining, forecasting and data optimization
Answer: c) Specialized software tools and applications for predictive analytics, data mining,
text mining, forecasting and data optimization.

72. What is the first challenge in Big Data analysis?

a) Breaking down datasilos to access all data an organization stores in different places and
often in different systems
b) Creating platforms that can pull in structured data as easily as unstructured data
c) Finding enough data to analyze
Answer: a) Breaking down datasilos to access all data an organization stores in different
places and often in different systems.

73. What is the second challenge in Big Data analysis?

a) Breaking down datasilos to access all data an organization stores in different places and
often in different systems
b) Creating platforms that can pull in structured data as easily as unstructured data
c) Finding enough data to analyze
Answer: b) Creating platforms that can pull in unstructured data as easily as structured data.
74. What is Deep Learning?
a) A subfield of machine learning that learns representations of data.
b) A type of computer hardware that is exceptionally effective at learning patterns.
c) A programming language used for data analysis.

Answer: a) A subfield of machine learning that learns representations of data.

75. What is the advantage of using learned features in Deep Learning?

a) They are over-specified and take a long time to design and validate.
b) They are easy to adapt and fast to learn.
c) They are effective for end-to-end joint system learning.

Answer: b) They are easy to adapt and fast to learn.

76. What type of information can Deep Learning represent?

a) Only visual information.
b) Only linguistic information.
c) World, visual, and linguistic information.

Answer: c) World, visual, and linguistic information.

77. What type of learning can Deep Learning perform?

a) Only supervised learning.
b) Only unsupervised learning.
c) Both supervised and unsupervised learning.

Answer: c) Both supervised and unsupervised learning.

78. What is the advantage of using Deep Learning techniques over other machine learning
techniques?
a) They require less training data.
b) They are easier to implement.
c) They outperform other techniques in certain domains.
Answer: c) They outperform other techniques in certain domains.

79. What is the process of training a Deep Learning network?

a) Forward it through the network, get predictions, and update the network weights.
b) Back propagate the errors, update the network weights, and get more accurate predictions.
c) Use error signal to change the weights and get more accurate predictions.

Answer: b) Back propagate the errors, update the network weights, and get more accurate
predictions.

80. How can gradient descent be used to optimize the weights in a Deep Learning network?
a) Adding a fraction of the gradient moves you towards the (local) minimum of the cost
function.
b) Subtracting a fraction of the gradient moves you towards the (local) minimum of the cost
function.
c) Multiplying a fraction of the gradient moves you towards the (local) minimum of the cost
function.

Answer: b) Subtracting a fraction of the gradient moves you towards the (local) minimum of
the cost function.
81. What is the purpose of non-linearities in Neural Networks?
a) To make the NN a linear function
b) To learn complex representations of data
c) To make the NN faster

Answer: b) To learn complex representations of data

82. What is the relationship between the number of layers and neurons in a NN and its ability
to approximate complex functions?
a) The more layers and neurons, the less complex functions it can approximate
b) The more layers and neurons, the more complex functions it can approximate
c) The number of layers and neurons does not affect its ability to approximate functions

Answer: b) The more layers and neurons, the more complex functions it can approximate

83. What is the activation function used in Sigmoid neurons?

a) Rectified Linear Unit (ReLU)
b) Tanh
c) Sigmoid

Answer: c) Sigmoid

84. What happens when a Sigmoid neuron saturates?

a) It fires at full capacity
b) It does not fire at all
c) It kills gradients and the NN barely learns

Answer: c) It kills gradients and the NN barely learns

85. What is the activation range of Tanh?

a) Between 0 and 1
b) Between -1 and 1
c) Thresholded at zero

Answer: b) Between -1 and 1

86. What is the advantage of using ReLU as an activation function over Sigmoid and Tanh?
a) ReLU is more expensive to implement
b) ReLU is less expressive
c) ReLU prevents the gradient vanishing problem and is less expensive to implement

Answer: c) ReLU prevents the gradient vanishing problem and is less expensive to
implement

87. What is overfitting in a NN?

a) When the NN fails to generalize to new examples
b) When the NN approximates complex functions too well
c) When the NN learns too slowly

Answer: a) When the NN fails to generalize to new examples

. What is big data?
a) A collection of data that is too large and complex for traditional data processing tools
b) A type of data that is only used by large corporations
c) A new kind of data that is being developed by scientists

Answer: a) A collection of data that is too large and complex for traditional data processing
tools

88. Which of the following is NOT a characteristic of big data?

a) Volume
b) Velocity
c) Variety
d) Value

Answer: d) Value

89. What is the term used to describe the process of extracting meaningful insights from big
data?
a) Data mining
b) Data warehousing
c) Data cleansing
d) Data validation

Answer: a) Data mining

90. Which of the following is an example of unstructured data?

a) Sales figures in a spreadsheet
b) Customer reviews on a website
c) Inventory levels in a database
d) Employee contact information in a directory

Answer: b) Customer reviews on a website

91. What is the purpose of Hadoop?

a) To store and manage big data
b) To analyze and visualize big data
c) To collect and process big data
d) To secure and protect big data

Answer: a) To store and manage big data

92. Which of the following is NOT a type of big data analytics?

a) Descriptive analytics
b) Predictive analytics
c) Prescriptive analytics
d) Reactive analytics

Answer: d) Reactive analytics

93. What is a data lake?

a) A type of swimming pool used to cool data servers
b) A repository for raw, unstructured data
c) A tool used to clean and organize big data
d) A method of compressing and storing big data

Answer: b) A repository for raw, unstructured data

94. Which of the following is a challenge of working with big data?

a) Lack of storage solutions
b) Limited processing power
c) Difficulty in finding skilled professionals
d) All of the above
Answer: d) All of the above

95. What is the difference between structured and unstructured data?

a) Structured data is organized and easy to analyze, while unstructured data is messy and hard
to work with
b) Structured data is textual, while unstructured data is numerical
c) Structured data is small in size, while unstructured data is large
d) Structured data is always collected from a single source, while unstructured data is
collected from multiple sources

Answer: a) Structured data is organized and easy to analyze, while unstructured data is messy
and hard to work with

96. What is the main benefit of using big data analytics?

a) Improved decision-making
b) Increased storage capacity
c) Reduced processing time
d) Enhanced security

Answer: a) Improved decision-making

97. Which of the following is NOT a type of machine learning algorithm?
a) Supervised learning
b) Unsupervised learning
c) Reinforcement learning
d) Deterministic learning

Answer: d) Deterministic learning

98. What is the purpose of a decision tree algorithm?

a) To classify data into categories
b) To predict numerical values
c) To group similar data points together
d) To identify anomalies in the data

Answer: a) To classify data into categories

99. Which of the following is a type of unsupervised learning algorithm?

a) Naive Bayes
b) K-nearest neighbors
c) K-means clustering
d) Linear regression

Answer: c) K-means clustering

100. What is the main difference between a neural network and a decision tree?
a) A decision tree is a type of neural network
b) A neural network can handle more complex data than a decision tree
c) A decision tree is better suited for regression problems than a neural network
d) A neural network is easier to explain than a decision tree

Answer: b) A neural network can handle more complex data than a decision tree

101. Which of the following is a type of reinforcement learning algorithm?

a) Linear regression
b) Random forest
c) Q-learning
d) Support vector machines

Answer: c) Q-learning

102. What is the purpose of a support vector machine algorithm?

a) To identify patterns in the data
b) To classify data into categories
c) To predict numerical values
d) To cluster similar data points together

Answer: b) To classify data into categories

103. Which of the following is NOT a step in the k-nearest neighbors algorithm?
a) Calculate the distance between the new data point and each existing data point
b) Choose the k closest data points
c) Assign the new data point to the category that appears most frequently among the k closest
data points
d) Fit a line or curve to the data points

Answer: d) Fit a line or curve to the data points

104. What is the main advantage of using a random forest algorithm?

a) It can handle missing data well
b) It is easy to explain and interpret
c) It is very fast and efficient
d) It can handle high-dimensional data well

Answer: d) It can handle high-dimensional data well

105. Which of the following is a type of deep learning algorithm?

a) Linear regression
b) K-means clustering
c) Convolutional neural networks
d) Decision trees

Answer: c) Convolutional neural networks

106. What is the purpose of a gradient descent algorithm?

a) To minimize the error between predicted and actual values
b) To maximize the accuracy of a classification model
c) To group similar data points together
d) To identify outliers in the data

Answer: a) To minimize the error between predicted and actual values

107. What is an artificial neuron?
a) A type of software that mimics the behavior of a biological neuron
b) A physical device that is used to model neural networks
c) A mathematical function that takes input values and produces an output value

Answer: c) A mathematical function that takes input values and produces an output value

108. What is the purpose of an activation function in an artificial neuron network?

a) To normalize the output values
b) To prevent overfitting
c) To introduce non-linearity into the network
d) To reduce the number of neurons required in the network

Answer: c) To introduce non-linearity into the network

109. What is the difference between a shallow neural network and a deep neural network?
a) A shallow neural network has fewer layers than a deep neural network
b) A shallow neural network has more layers than a deep neural network
c) A shallow neural network is only used for classification tasks, while a deep neural network
can be used for both classification and regression tasks
d) There is no difference between a shallow neural network and a deep neural network

Answer: a) A shallow neural network has fewer layers than a deep neural network

110. What is the purpose of backpropagation in a neural network?

a) To update the weights of the neurons based on the error in the output
b) To add more layers to the network
c) To remove unnecessary neurons from the network
d) To prevent overfitting

Answer: a) To update the weights of the neurons based on the error in the output
110. Which of the following is NOT a type of deep learning algorithm?
a) Convolutional neural networks
b) Recurrent neural networks
c) Decision trees
d) Generative adversarial networks

Answer: c) Decision trees

111. What is the purpose of a convolutional neural network?

a) To classify images and other visual data
b) To generate new data based on existing data
c) To predict numerical values
d) To cluster similar data points together

Answer: a) To classify images and other visual data

112. What is the purpose of a recurrent neural network?

a) To handle sequential data, such as time series or text data
b) To classify images and other visual data
c) To predict numerical values
d) To cluster similar data points together

Answer: a) To handle sequential data, such as time series or text data

113. What is the difference between supervised learning and unsupervised learning in deep
learning?
a) In supervised learning, the network is trained on labeled data, while in unsupervised
learning, the network is trained on unlabeled data
b) In supervised learning, the network is used for classification tasks, while in unsupervised
learning, the network is used for clustering tasks
c) In supervised learning, the network is more complex than in unsupervised learning
d) There is no difference between supervised learning and unsupervised learning in deep
learning

Answer: a) In supervised learning, the network is trained on labeled data, while in

unsupervised learning, the network is trained on unlabeled data

114. What is the purpose of a generative adversarial network?

a) To generate new data that is similar to existing data
b) To classify images and other visual data
c) To predict numerical values
d) To cluster similar data points together
Answer: a) To generate new data that is similar to existing data

115. What is the main advantage of using deep learning algorithms?

a) They can handle very large and complex datasets
b) They are very fast and efficient
c) They are very simple to implement and use
d) They can handle missing data well

Answer: a) They can handle very large and complex datasets

117. What is an artificial neuron network?

a) A set of interconnected neurons that process and transmit information
b) A type of computer hardware that mimics the structure of the human brain
c) A machine learning algorithm that uses decision trees to classify data
d) A software program that automatically labels images

Answer: a) A set of interconnected neurons that process and transmit information

118. What is an activation function in an artificial neuron network?

a) A function that determines the output of a neuron based on its inputs
b) A function that determines the learning rate of the network
c) A function that determines the number of layers in the network
d) A function that determines the number of neurons in each layer of the network

Answer: a) A function that determines the output of a neuron based on its inputs

119. What is backpropagation in an artificial neuron network?

a) A method for training the network by adjusting the weights of the connections between
neurons
b) A method for testing the network's accuracy on new data
c) A method for visualizing the network's structure
d) A method for reducing the number of neurons in the network

Answer: a) A method for training the network by adjusting the weights of the connections
between neurons

120. Which of the following is NOT a type of artificial neuron network?

a) Convolutional neural network
b) Recurrent neural network
c) Decision tree
d) Feedforward neural network

Answer: c) Decision tree

121. What is a feedforward neural network?
a) A type of artificial neuron network where the information flows in one direction, from
input to output
b) A type of artificial neuron network where the information flows in both directions, from
input to output and from output to input
c) A type of artificial neuron network where the information flows in a circular pattern
d) A type of artificial neuron network where the information flows randomly

Answer: a) A type of artificial neuron network where the information flows in one direction,
from input to output

122. What is a convolutional neural network?

a) A type of artificial neuron network commonly used in image recognition tasks
b) A type of artificial neuron network commonly used in text classification tasks
c) A type of artificial neuron network commonly used in speech recognition tasks
d) A type of artificial neuron network commonly used in time series forecasting tasks

Answer: a) A type of artificial neuron network commonly used in image recognition tasks

123. What is a recurrent neural network?

a) A type of artificial neuron network that can process sequences of inputs
b) A type of artificial neuron network that can only process individual inputs
c) A type of artificial neuron network that can only process images
d) A type of artificial neuron network that can only process text

Answer: a) A type of artificial neuron network that can process sequences of inputs

124. What is a generative adversarial network?

a) A type of artificial neuron network that can generate new data samples
b) A type of artificial neuron network that can classify data into categories
c) A type of artificial neuron network that can perform regression analysis
d) A type of artificial neuron network that can perform feature extraction

Answer: a) A type of artificial neuron network that can generate new data samples

125. What is deep learning?

a) A type of machine learning that uses artificial neuron networks with many layers
b) A type of machine learning that uses decision trees
c) A type of machine learning that uses rule-based systems
d) A type of machine learning that uses clustering algorithms

Answer: a) A type of machine learning that uses artificial neuron networks with many layers

126. What is transfer learning?

a) A technique for reusing a pre-trained artificial neuron network on a new task
b) A technique for training an artificial neuron network from scratch on every new task
c) A technique for combining multiple artificial neuron networks into a single network
d) A technique for visualizing the structure of an artificial neuron network

Answer: a) A technique for reusing a pre-trained artificial neuron network on a new task
1. What is a feed-forward network?
a) A type of neural network where data flows from the inputs to the outputs
b) A type of neural network where data flows from the outputs to the inputs
c) A type of neural network where data flows in a circular pattern
d) A type of neural network where data flows randomly

Answer: a) A type of neural network where data flows from the inputs to the outputs

127. Why is a feed-forward network acyclic?

a) Because there are no loops in the network
b) Because there are only forward connections between nodes
c) Because it has a memory
d) Because it can have arbitrary connections between nodes

Answer: a) Because there are no loops in the network

128. What is the state of a feed-forward network?

a) Fixed
b) Variable
c) Random
d) Non-existent

Answer: a) Fixed

129. What is a recurrent network?

a) A type of neural network that can have connections that go backward from output nodes to
input nodes
b) A type of neural network that can only have forward connections between nodes
c) A type of neural network that can only have backward connections between nodes
d) A type of neural network that can only have connections between adjacent nodes

Answer: a) A type of neural network that can have connections that go backward from output
nodes to input nodes

130. What is the difference between a feed-forward network and a recurrent network?
a) A feed-forward network does not have memory, while a recurrent network does
b) A feed-forward network has backward connections, while a recurrent network does not
c) A feed-forward network has arbitrary connections between nodes, while a recurrent
network does not
d) A feed-forward network has loops, while a recurrent network does not

Answer: a) A feed-forward network does not have memory, while a recurrent network does

131. What is the advantage of a recurrent network over a feed-forward network?

a) A recurrent network can have memory
b) A recurrent network is faster
c) A recurrent network is simpler
d) A recurrent network is more accurate

Answer: a) A recurrent network can have memory

132. What kind of connections can a recurrent network have?

a) Arbitrary connections between any nodes
b) Only forward connections between nodes
c) Only backward connections between nodes
d) Only connections between adjacent nodes

Answer: a) Arbitrary connections between any nodes

133. What is the internal state of a recurrent network?

a) It can alter as sets of input data are presented to it
b) It is fixed
c) It is random
d) It does not exist

Answer: a) It can alter as sets of input data are presented to it

134. What is the difference between a feed-forward network and a recurrent network in terms
of data flow?
a) In a feed-forward network, data flows from the inputs to the outputs, while in a recurrent
network, data can flow in any direction
b) In a feed-forward network, data flows from the outputs to the inputs, while in a recurrent
network, data flows from the inputs to the outputs
c) In a feed-forward network, data flows in a circular pattern, while in a recurrent network,
data flows randomly
d) In a feed-forward network, data flows randomly, while in a recurrent network, data flows
in a circular pattern

Answer: a) In a feed-forward network, data flows from the inputs to the outputs, while in a
recurrent network, data can flow in any direction
135. What is the main advantage of a recurrent network's memory?
a) It can process sequences of data
b) It is faster than a feed-forward network
c) It is more accurate than a feed-forward network
d) It can handle larger datasets than a feed-forward network

Answer: a) It can process sequences of data

136. What is an artificial neural network?
a. A pool of simple processing units that communicate by sending signals to each other over a
large number of weighted connections.
b. A type of computer software that can learn from data.
c. A type of machine learning algorithm that can only process structured data.
d. None of the above.

Answer: a.

137. What is the main function of an artificial neural network?

a. To process data similar to the human brain.
b. To make decisions or take actions based on data.
c. To develop machines with human-like imagination and reasoning power.
d. None of the above.

Answer: a.

138. What is the difference between human brain and artificial neural network?
a. Human brain is more powerful than artificial neural network.
b. Artificial neural network is more powerful than human brain.
c. Both are equally powerful.
d. None of the above.

Answer: a.

139. What are the processing units in an artificial neural network?

a. Neurons.
b. Synapses.
c. Axons.
d. None of the above.

Answer: a.

140. How do the processing units in an artificial neural network communicate?

a. By sending signals to each other over a large number of weighted connections.
b. By processing data in parallel.
c. By using algorithms to make decisions.
d. None of the above.

Answer: a.

150. What is the purpose of weighted connections in an artificial neural network?

a. To allow the processing units to communicate with each other.
b. To regulate the flow of data through the network.
c. To increase the processing power of the network.
d. None of the above.

Answer: a.

151. What is the benefit of using artificial neural networks in machine learning?
a. They help machines learn from the tasks they perform.
b. They can process any type of data, regardless of its structure.
c. They can make decisions faster than humans.
d. None of the above.

Answer: a.

152. What is the limitation of artificial neural networks?

a. They cannot process unstructured data.
b. They require large amounts of data to learn effectively.
c. They can only make decisions based on the data they have been trained on.
d. None of the above.

Answer: b.

153. What is the role of ANNs in developing AI?

a. ANNs are a key component in developing AI.
b. ANNs are not necessary for developing AI.
c. ANNs are only useful for developing certain types of AI.
d. None of the above.

Answer: a.

154. What is the future of ANNs?

a. ANNs will continue to evolve and improve.
b. ANNs will eventually replace human intelligence.
c. ANNs will become obsolete as new technology emerges.
d. None of the above.

Answer: a.
155. What are the two basic reasons for building artificial neural networks?
a. Technical viewpoint and biological viewpoint.
b. Technical viewpoint and computational viewpoint.
c. Computational viewpoint and biological viewpoint.
d. None of the above.

Answer: a.

156. What kind of problems require massively parallel and adaptive processing?
a. Problems related to character recognition.
b. Problems related to predicting future states of a system.
c. Both a and b.
d. None of the above.

Answer: c.

157. How is information transmitted in biological systems?

a. As a series of electric impulses.
b. As a series of chemical reactions.
c. As a series of mechanical vibrations.
d. None of the above.

Answer: a.

158. What is the receptive field of a neuron?

a. The area where a neuron receives its information from other neurons.
b. The area where a neuron sends its information to other neurons.
c. The area where a neuron processes information.
d. None of the above.

Answer: a.

159. What is an artificial neural network?

a. A hardware implementation or a computer program which strives to simulate the
information processing capabilities of its biological exemplar.
b. A type of machine learning algorithm that can only process structured data.
c. A type of computer software that can learn from data.
d. None of the above.

Answer: a.

160. What are artificial neurons?

a. Simplified models of their biological counterparts.
b. Exact replicas of biological neurons.
c. Modified versions of biological neurons.
d. None of the above.

Answer: a.

161. What is the purpose of an artificial neural network?

a. To simulate the information processing capabilities of its biological exemplar.
b. To process data in a way that is similar to the human brain.
c. To solve problems by constructing software that works like our brains.
d. All of the above.

Answer: d.

162. What is the strength of the signal that reaches the next neuron dependent on?
a. The amount of neurotransmitter.
b. The type of neurotransmitter.
c. The frequency and phase of the electric impulses.
d. None of the above.

Answer: a.

163. How many neurons are typically found in a human brain?

a. 10,000.
b. 1 million.
c. 10 billion.
d. None of the above.

Answer: c.

164. What is the purpose of neurotransmitters in biological systems?

a. To transmit information from one neuron to the next.
b. To regulate the strength of the signal that reaches the next neuron.
c. To encode information as a series of electric impulses.
d. None of the above.

Answer: a.
165. What is the function of dendrites in artificial neural networks?
a. They represent nodes.
b. They represent weights.
c. They represent inputs.
d. None of the above.

Answer: c.

166. What is the function of the cell nucleus in artificial neural networks?
a. They represent nodes.
b. They represent weights.
c. They represent inputs.
d. None of the above.

Answer: a.

167. What is the function of synapses in artificial neural networks?

a. They represent nodes.
b. They represent weights.
c. They represent inputs.
d. None of the above.

Answer: b.

168. What is the function of axons in artificial neural networks?

a. They represent nodes.
b. They represent weights.
c. They represent inputs.
d. They represent outputs.

Answer: d.

169. What is the perceptron?

a. A type of artificial neuron.
b. A type of artificial neural network.
c. A type of computer program.
d. None of the above.

Answer: a.

170. What is the step function used by a perceptron?

a. It returns +1 if the weighted sum of the inputs is greater than a threshold, t, and 0 if X is
less than or equal to t.
b. It returns the weighted sum of the inputs.
c. It returns the product of the inputs.
d. None of the above.

Answer: a.

171. What is the purpose of the variables w0 and x0 in the perceptron?

a. To represent the input.
b. To represent the weights.
c. To represent the threshold.
d. None of the above.

Answer: c.

172. What is the purpose of backpropagation in neural networks?

a. To minimize the error between the layer's response and the actual data.
b. To maximize the error between the layer's response and the actual data.
c. To adjust the weights randomly.
d. None of the above.

Answer: a.

173. What is the purpose of hidden layers in neural networks?

a. To classify inputs into one of two categories.
b. To represent true and false.
c. To increase the power of the neural network.
d. None of the above.

Answer: c.