AI For Engineering (KMC101/KMC201)- MCQ

Q:1. A.M. Turing developed a technique for determining whether a computer could or could not
demonstrate the artificial Intelligence, Presently, this technique is called __________

1.Turing Test

2.Algorithm

3.Boolean Algebra

4.Logarithm

Solution- 1. Turing Test

Reason 1- The test is known as after Turing , the founding father of the Turing Test and an
English scientist , cryptanalyst, mathematician and theoretical biologist.

Reason 2- The test is understood as after Turing , the founding father of the Turing Test and an
English scientist , cryptanalyst, mathematician and theoretical biologist.

Reason 3- The test is named after Alan Turing , the founding father of the Turing Test and an English
scientist , cryptanalyst, mathematician and theoretical biologist.

Q:2. Knowledge based systems comprises of:

1.DENDRAL

2.MYCIN

3.PROSPECTOR

4.All the above

Solution- 1. DENDRAL

Reason 1- The DENDRAL programs were knowledge-driven, within the sense of today’s expert
systems, with the knowledge principle–that knowledge is power-first articulated within the context
of DENDRAL

Reason 2- The DENDRAL programs are knowledge-driven, and within the sense of today’s expert
systems, with the knowledge principle–that the knowledge is the power-first articulated within the
context of DENDRAL

Q:3. Weak AI is

1.A set of computer programs that produce output that would be considered to reflect intelligence if
it were generated by humans.

2.Useful for testing hypothesis about minds, but would not actually be minds

3.The embodiment of human intellectual capabilities within a computer.

4.None of the above

Solution- 2.Useful for testing hypothesis about minds, but would not actually be minds

Reason 1- Weak AI lacks human consciousness, although it’s going to be ready to simulate
it sometimes .

Reason 2- Weak AI lacks human consciousness, although it may be able to simulate it at times.

Reason 3- Weak AI lacks human consciousness, although it’s getting to be able to simulate it
sometimes.

Q:4. Strong AI is

1.An AI system with generalized human cognitive abilities.

2.Also called as narrow AI.

3.All actions are preprogrammed by human

4.None of the above

Solution- 1.An AI system with generalized human cognitive abilities.

Reason 1- Strong AI include the ability to reason, solve puzzles, make judgments, plan, learn, and
communicate. It should also have consciousness, objective thoughts, self-awareness, sentience, and
sapience. Strong AI is also called True Intelligence or Artificial General Intelligence (AGI)

Reason 2- Strong AI include the power to reason, solve puzzles, make judgments, plan, learn, and
communicate. It should even have consciousness, objective thoughts, self-awareness, sentience, and
sapience. Strong AI is additionally called True Intelligence or Artificial General Intelligence (AGI)

Reason 3- Strong AI include the facility to reason, solve puzzles, make judgments, plan, learn, and
communicate. It should even have consciousness, objective thoughts, self-awareness, sentience, and
sapience. Strong AI is additionally called True Intelligence or Artificial General Intelligence (AGI)

Q:5. Which of the following could be the approaches to Artificial Intelligence?

1.Strong AI

2.Swarm Intelligence

3.Computational Intelligence

4.All the above

Solution- 4. All of the Above

Reason 1- Artificial Intelligence comprises of the following abilities such as Strong A.I, Swarm
Intelligence and computational Intelligence.

Reason 2- Artificial Intelligence comprises of the subsequent abilities like Strong A.I, Swarm
Intelligence and computational Intelligence.
Reason 3- Artificial Intelligence comprises of the next abilities like Strong A.I, Swarm Intelligence and
computational Intelligence.

Q:6. AI vs. Human Brain

1.Humans use content memory and thinking whereas, robots are using built-in instructions,
designed by scientists.

2.Artificial intelligence cannot beat human intelligence at all

3.The field of Artificial intelligence limits on designing machines that can mimic human behavior.

4.None of the above

Solution- 3.The field of Artificial intelligence limits on designing machines that can mimic human
behavior.

Reason 1- Because these machines are harder to implement and requires tons of memory usage and
processing power.

Reason 2- Because these machines are harder to implement and requires a lot of memory usage and
processing power.

Reason 3- because these machines are harder to implement and requires plenty of memory usage
and processing power.

Q:7. Which of the following is not a stage of AI?

1.Cognitive analytics

2.Predictive analytics

3.Diagnostic analytics

4.None of the above

Solution- 1. Cognitive Analytics

Reason 1- Cognitive Computing tries to replicate how humans would solve problems while AI seeks
to create new ways to solve problems that can potentially be better than humans.

Reason 2- Cognitive Computing tries to duplicate how humans would solve problems while AI seeks
to make new ways to unravel problems which will potentially be better than humans.

Reason 3- Cognitive Computing tries to duplicate how humans would solve problems while AI seeks
to form new ways to unravel problems which can potentially be better than humans.

Q:8. Steps to process the command are:

1.Trigger word detection

2.Speech recognition
3.Intent recognition

4.All the above

Solution 4- All of the above

Reason 1- All of the command that are written within the above option are the steps to process the
commands.

Reason 2- All of the command that are written in the above option are the steps to process the
commands.

Q:9. Which of the following is not an application of AI?

1.Pattern recognition

2.Crop prediction

3.Digital assistant

4.Fund transfer

Solution- 4.Fund transfer

Reason 1- As we know pattern recognition, crop prediction, digital assistant all come under the
application of A.I but fund transfer is the process that is done using pre programmed instructions.

Reason 2- As we all know pattern recognition, crop prediction, digital assistant all come under the
appliance of A.I but fund transfer is that the process that’s done using pre programmed instructions.

Q:11. A method of programming a computer to exhibit human intelligence is called modeling

or…………………

1.simulation

2.cognitization

3.duplication

4.psychic amelioration

Solution- 1.Simulation

Reason 1- because the production of a computer model of something, especially for the purpose of
study is known as Simulation.

Reason 2- because the assembly of a computer model of something, especially for the aim of study
is understood as Simulation.

Q:1. What is an advantage of Artificial Intelligence?

1.Potential for misuse

2.Highly dependent on machines

3.Requires Supervision

4.Rational Decision Maker

Solution- 4- Rational Decision Maker

Reason 1- This is because Rational choice theory in A.I. states that it rely on rational calculations to
make rational choices that result in outcomes aligned with their own best interests.

Reason 2- This is because Rational choice theory during a .I. states that it believe rational
calculations to form rational choices that end in outcomes aligned with their own best interests.

Reason 3- This is because Rational choice theory during artificial intelligence states that it believe
rational calculations to make rational choices that end in outcomes aligned with their own best
interests.

Q:12. Who is known as the “Father of AI”?

1.Fisher Ada

2.Alan Turing

3.John McCarthy

4.Allen Newell

Solution- 3.John McCarthy

Reason 1- John McCarthy, widely recognized as the father of Artificial Intelligence due to his
astounding contribution in the field of Computer Science and AI.

Reason 2- John McCarthy, widely known because the father of AI thanks to his astounding
contribution within the field of computing and AI.

Reason 3- John McCarthy, widely known because the father of Artificial intelligence because of his
astounding contribution within the sector of computing and AI.

Q:13. Which of the following is not a branch of Artificial Intelligence?

1.Expert systems

2.Robotics

3.Natural language Processing

4.None of the above

Solution- 4.None of the above

Reason 1- This is because Expert System, Robotics and Natural Language processing all have their
roles in Artificial Intelligence. There are 6 branches of A.I and these 3 above come under these
branches.
Reason 2- This is because Expert System, Robotics and NLP all have their roles in AI . There are 6
branches of A.I and these 3 above come under these branches.

Reason 3- This is because Expert System, Robotics and NLP processing have their roles in AI . There
are six branches of A.I and these three above come under these branches.

Q:14. Which of the following is not an application of Unsupervised Learning?

1.Document clustering

2.Speech recognition

3.Image compression

4.Association analysis

Solution- 4. Speech Recognition

Reason 1- This is the kind of application where you teach the algorithm about your voice and it will
be able to recognize you. The most well-known real-world applications are virtual assistants such as
Google Assistant and Siri, which will wake up to the keyword with your voice only. That is the reason
it comes under supervised learning.

Reason 2-This is the type of application where you teach the algorithm about your voice and it’ll be
ready to recognize you. the foremost well-known real-world applications are virtual assistants like
Google Assistant and Siri, which can awaken to the keyword together with your voice only. that’s the
rationale it comes under supervised learning.

Reason 3-This is the type of application where you teach the algorithm about your voice and it’ll be
ready to recognize you. the foremost well-known real-world applications are virtual assistants like
Google Assistant and Siri, which can awaken to the keyword in conjunction with your voice only.
that’s the rationale it comes under supervised learning.

Q:15. The multi-armed bandit problem is a generalized use case for-

1.Reinforcement learning

2.Supervised learning

3.Unsupervised learning

4.All the above

Solution- 1.Reinforcement learning

Reason 1- Multi–Arm Bandit is a classic reinforcement learning problem, in which a player is facing
with k slot machines or bandits, each with a different reward distribution, and the player is trying to
maximise his cumulative reward based on trials.

Reason 2- Multi-Arm Bandit may be a classic reinforcement learning problem, during which a player
is facing with k slot machines or bandits, each with a special reward distribution, and therefore the
player is trying to maximise his cumulative reward supported trials.
Reason 3- Multi-Arm Bandit could also be a classic reinforcement learning problem, during which a
player is facing with k slot machines or bandits, each with a special reward distribution, and thus the
player is trying to maximise his cumulative reward supported trials.

Q:16. Why IOT now?

1.Electronic companies are building Wi-Fi and cellular wireless connectivity into a wide range of
devices.

2.Mobile data coverage has improved significantly

3.The size and cost of wireless radios has dropped

4.All the above

Solution- 4.All the above

Reason 1- All of the above options that have provided in this question are correct and helping IOT to
evolve in the future.

Reason 2- All of the above options that have provided during this question are correct and helping
IOT to evolve within the future.

Q:17. Scalability of IoT means:

1.Expandable/reducible in terms of scale or size.

2.Measurable

3.Increasing/decreasing monetary costs.

4.All of these.

Solution- 4.All of these.

Reason 1- This is because the IOT is expandable/reducible in size or to scale. It is also measurable
and its cost can be increased or decreased.

Reason 2- This is because the IOT is expandable/reducible in size or to scale. it’s also measurable and
its cost are often increased or decreased.

Q:18. Which of the following statement is incorrect for AI?

1.Humans can not evolve as soon as AI evolves to control and handle it

2.The community working towards safe and beneficial superintelligence has grown worldwide.

3.It is typically managed by a peer-to-peer network working simultaneously together to solve

complex mathematical problems in order to validate new blocks

4.AI is the new electricity

Solution- 1.Humans can not evolve as soon as AI evolves to control and handle it
Reason 1- We cannot say humans cannot evolve to control and handle it because humans are
working on it regularly to make an A.I that will reduce their workload and time to complete a task,
which they will adapt later.

Reason 2- We cannot say humans cannot evolve to regulate and handle it because humans are
performing on it regularly to form an A.I which will reduce their workload and time to finish a task,
which they’re going to adapt later.

Reason 3- We cannot say humans cannot evolve to manage and handle it because humans are
working on it regularly to make an A.I which can reduce their workload and time to end a task, which
they go to adapt later.

Q:19. Overfitting

1.The model remembers a huge number of examples instead of learning to notice features and may
fail to predict future observations reliably.

2.Occurs when a statistical model cannot adequately capture the underlying structure of the data.

3.Occurs if the model or algorithm shows low variance but high bias

4.None of the above

Solution- 1.The model remembers a huge number of examples instead of learning to notice
features and may fail to predict future observations reliably.

Reason 1- Overfitting is “the production of an analysis that corresponds too closely or exactly to a
particular set of data, and may therefore fail to fit additional data or predict future observations
reliably”

Reason 2- Overfitting is “the production of an analysis that corresponds too closely or exactly to a
specific set of knowledge , and should therefore fail to suit additional data or predict future
observations reliably”

Q:20. AI in security will:

1.Not detect threats based on application behavior and a whole network’s activity.

2.Not be able to identify and stop cyber threats with less human intervention than is typically
expected or needed with traditional security approaches.

3.Not detect when new threats are imminent

4.Not replace a security analyst’s insights or understanding of the field.

Solution- 4. Not replace a security analyst’s insights or understanding of the field.

Reason 1- This is because we do not have A.I’s that can understand and gathered field intelligence
without human intervention.

Reason 2- This is because we don’t have A.I’s which will understand and gathered field intelligence
without human intervention.
Q:21. The first usage of Data came in:

1.1640

2.1954

3.1946

4.1940

Solution- 3. 1946

Reason- The first English use of the word “data” is from the 1640s. The word “data” was first
used to mean “transmissible and storable computer information” in 1946.

Q:22. DIKW:

1.Stands for Data, Information, Knowledge, Wisdom

2.In 1994 Nathan Shedroff presented the DIKW hierarchy in an information design context

3.In this context data is considered as symbols representing signals

4.All the above statements are correct

Solution- 1.Stands for Data, Information, Knowledge, Wisdom

Reason 1- Knowledge Pyramid, Wisdom Hierarchy and Information Hierarchy are some of the names
referring to the popular representation of the relationships between data, information, knowledge
and wisdom in the Data, Information, Knowledge, Wisdom (DIKW) Pyramid.

Reason 2- Knowledge Pyramid, Wisdom Hierarchy and knowledge Hierarchy are a number of the
names pertaining to the favored representation of the relationships between data, information,
knowledge and wisdom within the Data, Information, Knowledge, Wisdom (DIKW) Pyramid.

Q:23. Classification data type which is not on the basis of measurement:

1.Ratio data

2.Ordinal data

3.Boolean data (True/False)

4.Interval data

Solution- 3. Boolean data (True/False)

Reason- There are four levels of data measurements in classification data type: Nominal, Ordinal,
Interval, and Ratio.

Q:24. Not a case of Qualitative vs Quantitative data:

1.Category vs Number
2.Observed vs Measured

3.Smell vs Height

4.Volume vs Color

Solution- 4.Volume vs Color

Reason 1- Quantitative data can be counted, measured, and expressed using numbers. Qualitative
data is descriptive and conceptual. Qualitative data can be categorized based on traits and
characteristics.

Reason 2- Quantitative data are often counted, measured, and expressed using numbers. Qualitative
data is descriptive and conceptual. Qualitative data are often categorized supported traits and
characteristics.

Q:25. User driven approach is

1.Data Mining

2.Deep Learning

3.OLTP

4.Machine Learning

Solution- 3.OLTP

Reason 1- Current data warehouse development methods. can fall within three basic groups: data –
driven, goal-driven and user–driven. Implementation strategies.

Reason 2- Current data warehouse development methods. can fall within three basic groups: data -
driven, goal-driven and user-driven. Implementation strategies.

Q:26. Physical storage of data:

1.CD-ROM

2.Distributed database

3.Cloud storage

4.None of the above

Solution- 1.CD-ROM

Reason 1- Physical (non-electronic) data may be stored in a variety of forms including photographs,
film, optical media (e.g. CDs & DVDs), magnetic media (e.g. audio and video tapes or
computer storage devices), artworks, paper documents or computer printouts.

Reason 2- Physical (non-electronic) data could also be stored during a sort of forms including
photographs, film, optical media (e.g. CDs & DVDs), magnetic media (e.g. audio and video tapes or
memory devices), artworks, paper documents or computer printouts.
Q:27. Which of the following statement is true for Data Warehouse?

1.It is semi-structured and raw

2.It is less agile with fixed configuration

3.It is designed for low-cost storage

4.All the above

Solution- 2.It is less agile with fixed configuration

Reason 1- A data warehouse is a highly structured data bank, with a fixed configuration and little
agility. Changing the structure isn’t too difficult, at least technically, but doing so is time consuming
when you account for all the business processes that are already tied to the warehouse.

Reason 2- A data warehouse may be a highly structured data bank, with a hard and fast
configuration and tiny agility. Changing the structure isn’t too difficult, a minimum of technically, but
doing so is time consuming once you account for all the business processes that are already tied to
the warehouse.

Q:28. Importance of data:

1.It helps to analyze and visualize the performance

2.It helps to recommend correct options to the customers

3.It helps to solve complex problems

4.All the above

Solution- 4.All the above

Reason- All of the above steps are true in case of data. It helps us to in all the cases.

Q:29. Choose an incorrect statement:

1.ETL stands for Extraction, Transformation, Loading into repository.

2.Data cleaning is very important in data preparation.

3.Removal of outliers and smoothing of data is required to prepare data for further processing.

4.Data needs to be normalize.

Solution- 4.Data needs to be normalize.

Reason 1- Similarly, the goal of normalization is to change the values of numeric columns in the
dataset to a common scale, without distorting differences in the ranges of values. For machine
learning, every dataset does not require normalization. It is required only when features have
different ranges.
Reason 2- Similarly, the goal of normalization is to vary the values of numeric columns within the
dataset to a standard scale, without distorting differences within the ranges of values. For machine
learning, every dataset doesn’t require normalization. it’s required only features have different
ranges.

Q:30. Data visualization tools are:

1.Pie chart

2.Histogram

3.Scatter Plot

4.All the above

Solution- 4.All the above

Reason 1- Because data visualization tools are used to represent data in pictorial form. and all the
option above are used to represent data in visual form.

Reason 2- Because data visualization tools are wont to represent data in pictorial form. and every
one the choice above are wont to represent data in visual form.

Q:31. In Supervised Learning:

1.Input data is called training data and has a known label.

2.It can solve the classification and regression problems.

3.The training process continues until model achieves desired accuracy

4.All the above statements are true.

Solution- 4.All the above statements are true.

Reason 1- All the three 1,2, and 3 statements are true. As in supervised learning input data is called
training data and has a known label. This is also used to solve the classification and regression
problems. This training process continues until model achieves desired accuracy.

Reason 2- All the three 1,2, and three statements are true. As in supervised learning input file is
named training data and features a known label. this is often also wont to solve the classification
and regression problems. This training process continues until model achieves desired accuracy.

Q:32. In Unsupervised Learning, the incorrect statements are:

1.It organize data by similarity.

2.Input data know about results

3.It can solve problem of dimension reduction.

4.None of the above

Solution- 4.None of the above

Reason 1- This is because it organizes data by similarity and the input data know about the result. It
is also used to solve dimensionality problem.

Reason 2- This is because it organizes data by similarity and therefore the input file realize the result.
it’s also wont to solve dimensionality problem.

Q:33. Data Visualization is:

1.Used to communicate information clearly and efficiently to users by the usage of information
graphics such as tables and charts.

2.Helps users in analyzing a large amount of data in a simpler way.

3.Makes complex data more accessible, understandable, and usable.

4.All of the above

Solution- 4. All of the above

Reason 1- All the reason that has been provided above are true. because Data visualization is a
process of representing data into pictorial or graphical form.

Reason 2- All the rationale that has been provided above are true. because Data visualization may
be a process of representing data into pictorial or graphical form.

Q:34. Data Visualization tool that can be used for displaying hierarchical data:

1.Histogram

2.Treemap

3.Scatter plot

4.Pie chart

Solution- 2.Treemap

Reason 1- Treemaps are visualizations for hierarchical data. They are made of a series of nested
rectangles of sizes proportional to the corresponding data value.

Reason 2- Treemaps are visualizations for hierarchical data. they’re made from a series of nested
rectangles of sizes proportional to the corresponding data value.

Q:35. Which of the following is a Regression problem?

1.Weather forecasting

2.Spam/Not-Spam emails categorization

3.Sentiment analysis

4.Fraud detection
Solution- 4. Fraud Detection

Reason 1- In order to effectively test, detect, validate, correct error and monitor control systems
against fraudulent activities, businesses entities and organizations rely on specialized data analytics
techniques such as data mining, data matching, sounds like function, Regression analysis, Clustering
analysis and Gap.

Reason 2- In order to effectively test, detect, validate, correct error and monitor control systems
against fraudulent activities, businesses entities and organizations believe specialized data analytics
techniques like data processing , data matching, seems like function, multivariate analysis ,
Clustering analysis and Gap.

Q:36. Which of the following is a Classification problem?

1.Estimating the price of house

2.Credit/loan approval

3.Recommender system

4.Predicts the number of items which a consumer will probably purchase

Solution- 2. Credit/loan approval

Reason 1- Classification algorithms work by predicting the best group to which a data point belongs
to by learning from labeled observations. It uses a set of input features for the learning process.
Classification algorithms are good for grouping data that are never seen before into their various
groupings and are therefore extensively used in machine learning tasks.

Reason 2- Classification algorithms work by predicting the simplest group to which a knowledge
point belongs to by learning from labeled observations. It uses a group of input features for the
training process. Classification algorithms are good for grouping data that are never seen before into
their various groupings and are therefore extensively utilized in machine learning tasks.

Q:37. Decision tree:

1.Belongs to a family of unsupervised learning algorithms

2.Consider all attributes to split at each node, starting from the root node

3.Create a model that can be used to predict the class or value of the target variable by learning
simple decision rules inferred from training data

4.All the above

Solution- 3.Create a model that can be used to predict the class or value of the target variable by
learning simple decision rules inferred from training data

Reason 1- because it is used to create a model that can be used to predict the class or value of the
target variable by learning simple decision rules inferred from training data

Reason 2- because it’s wont to create a model which will be wont to predict the category or value of
the target variable by learning simple decision rules inferred from training data
Q:38. Bayesian Classifier:

1.Connects the degree of belief in a hypothesis before and after accounting for evidence

2.Uses conditional and marginal probability

3.Performance can be estimated using accuracy, precision, recall

4.All the above

Solution- 4.All the above

Reason 1- This is because Bayesian Classifier uses conditional and marginal probability. Also it
connects the degree of belief in a hypothesis before and after accounting for evidence. And its
performance can be estimated using accuracy, precision, recall. That is the reason all the above
options are true.

Reason 2- This is because Bayesian Classifier uses conditional and marginal probability. Also it
connects the degree of belief during a hypothesis before and after accounting for evidence. And its
performance are often estimated using accuracy, precision, recall. that’s the rationale all the above
options are true.

Q:39. When two clusters have a parent-child relationship then it is called as:

1.K-means clustering

2.Fuzzy c-means clustering

3.Hierarchical clustering

4.Density based clustering

Solution- 3.Hierarchical clustering

Reason 1- When two clusters have parent-child relationship or tree like structure then it is called as
Hierarchical Clustering.

Q:40. Recommender system is an example of:

1.Clustering

2.Supervised learning

3.Reinforcement learning

4.Regression

Solution- 2. Supervised learning

Reason 1- The previous recommendation algorithms are rather simple and are appropriate for small
systems. Until this moment, we considered a recommendation problem as a supervised machine
learning task.
Q:41. The possible features of a text corpus in NLP

1.Count of the word

2.Identifying stop words

3.Predicting parts of Speech

4.All the above

Solution- 4.All the above

Reason 1- All of the above options are true.

Reason 2- the above options are true.

Q:42. Normalization techniques in NLP

a. Lemmatization
b. Bag of words
c. Stemming
d. Named entity recognition

1.a,b

2.a,c

3.d,b

4.b,c

Solution- 2. a,c

Reason 1- In NLP a highly overlooked preprocessing step is text normalization. and Lemmatization
on the surface is very similar to stemming, where the goal is to remove inflections and map a word
to its root form. So these two are the normalization techniques in NLP.

Reason 2- In NLP a highly overlooked preprocessing step is text normalization. and Lemmatization
on the surface is extremely almost like stemming, where the goal is to get rid of inflections and map
a word to its root form. So these two are the normalization techniques in NLP.

Q:43. NLP Use cases

a. Text summarization
b. Object detection
c. Sentiment analysis
d. Chatbots

1.b,c,d

2.a,b,d
3.a, c, d

4.a,b,c

Solution- 3. a, c, d

Reason 1- Text summarization, Sentiment analysis and chatbots all uses natural language processing.

Q:44. Speech recognition

1.It is a way of encoding and decoding signals

2.It is coupled with AI as deep learning models

3.Both acoustic modeling and language modeling are important parts of modern statistically-based
speech recognition algorithms.

4.All the above

Solution- 4. All the above

Reason 1- This is because in speech recognition, encoding and decoding signals is done and it is also
coupled with AI as a deep learning model. and both the acoustic modeling and language modeling
are important parts of modern statistically-based speech recognition algorithms.

Reason 2- This is because in speech recognition, encoding and decoding signals is completed and it’s
also including AI as a deep learning model. and both the acoustic modeling and language modeling
are important parts of recent statistically-based speech recognition algorithms.

Q:45. Choose an incorrect statement in context of speech recognition

1.In 1952, three Bell Labs researchers built a system called “Audrey”

2.Modern general-purpose speech recognition systems are based on Hidden Markov Models

3.It can identify objects, people, places, and actions in images

4.None of the above

Solution 3- It can identify objects, people, places, and actions in images

Reason 1- This is because speech recognition do not identify objects, people, places, and actions in
images. This identification is done by object recognition and face recognition.

Reason 2- This is because speech recognition don’t identify objects, people, places, and actions in
images. This identification is completed by visual perception and face recognition.

Q:46. Natural Language Understanding (NLU)

a. It is the ability of machines to understand the human language

b. It is a branch of Natural Language Processing
c. Natural-language understanding is considered an AI-hard problem.
d. None of the above

1.a,b,c

2.a,c,d

3.b,a,d

4.b,c,d

Solution- 1.a,b,c

Reason 1- NLU is a branch of NLP and it is considered an AI hard problem. It is the ability of machines
to understand the human language

Q:47. Speech recognition steps include

1.Feature extraction

2.Spectrum analysis

3.Preprocessing of input signals

4.All the above

Solution- 4. All the above

Reason 1- Speech recognition process takes place in three main steps which are
acoustic processing, feature extraction and classification/recognition. It also include spectrum
analysis and preprocessing of input signals.

Reason 2- Speech recognition process takes place in 3 main steps which are acoustic processing,
feature extraction and classification/recognition. It also include spectrum and preprocessing of input
signals.

Q:48. The interpretation capabilities of a language-understanding system depend on

1.The semantic Theory

2.The syntactic theory

3.Both a and b

4.None of the above

Solution- 1. The semantic Theory

Reason 1- The interpretation capabilities of a language-understanding system depend on the

semantic theory it uses. Semantic parsers convert natural-language texts into formal meaning
representations.

Q:49. Applications of NLU

a. Automated reasoning
b. Machine translation
c. Network congestion control
d. All the above

1.c,d

2.b,c

3.d,a

4.a,b

Solution- 4.a,b

Reason 1- Before a computer can process unstructured text into a machine-readable format, first
machines need to understand the peculiarities of the human language. It gives machines a form of
reasoning or logic, and allows them to infer new facts by deduction. Simply put, using previously
gathered and analyzed information, computer programs are able to generate conclusions.

Reason 2- Before a computer can process unstructured text into a machine-readable format, first
machines got to understand the peculiarities of the human language. It gives machines a sort of
reasoning or logic, and allows them to infer new facts by deduction. Simply put, using previously
gathered and analyzed information, computer programs are ready to generate conclusions.

Q:50. Methods used in speech recognition systems are

1.Hidden Markov Model (HMM)

2.Neural Networks

3.Both a and b

4.None of the above

Solution- 3. Both a and b

Reason 1- at CMU, Raj Reddy’s students James Baker and Janet M. Baker began using the Hidden
Markov Model (HMM) for speech recognition. speech recognition was still dominated by traditional
approaches such as Hidden Markov Models combined with feedforward artificial neural networks.

Q:51. Choose an incorrect statement in context to Natural Language Generation (NLG)

1.Transforms structured data into natural language

2.Markov chains can be used for generating natural language

3.It converts a text into structured data

4.None of the above

Solution- 1.Transforms structured data into natural language

Reason 1- Natural Language Generation (NLG), a subcategory of Natural Language Processing
(NLP), is a software process that automatically transforms structured data into human-readable
text.

Q:52. In natural-language understanding, the system needs to disambiguate the input sentence to
produce the machine representation language, in NLG the system needs to make decisions about
how to put a concept into words

1.True

2.False

Solution- 1. True

Reason 1- In natural language understanding the system needs to disambiguate the input sentence
to produce the machine representation language, whereas in Natural Language Generation the
system needs to make decisions about how to put a concept into words.

Q:53. Applications of Natural Language Generation

a. Smartphone
b. Analysis of business intelligence
c. IOT devices
d. Chatbots

1.a, b, c, d

2.a,c,d

3.b,c

4.a,b,c

Solution- 1.a, b, c, d

Reason 1- All the above application given in the option are true and these are the applications of
NLG.

Q:54. Choose correct options

a. NLU takes up the understanding of the data based on grammar, the context in which it was said
and decide on intent and entities.
b. NLP converts a text into structured data.
c. NLG generates a text based on structured data.
d. None of the above

1.b,c,d

2.a,c
3.a,b,c

4.a,b,d

Solution- 3.a,b,c

Reason 1-

The way these three of them work hand in hand are given as- NLU takes up the understanding of the
data based on grammar, the context in which it was said and decide on intent and entities. NLP will
convert the text into structured data. NLG generates text generated based on structured data.

Q:55. Chatbots

a. Can be used for E-commerce

b. Can be used to solve people’s travel related problems
c. Need not to pass the industry standard Turing test at any level
d. Require a large amount of conversational data to train

1.b,c,d

2.a,c

3.a, b, c

4.a,b,d

Solution- 4.a,b,d

Reason 1- Chatbots can be used for E-commerce and to solve people’s travel related problems but it
require a large amount of conversational data to train.

Q:56. Machine translation

a. Is the process of using computer programs to translate a text/speech from one natural language
to another relevant to context
b. It has the ability to translate in many languages
c. It is required for web content and web page translation
d. None of the above

1.a,b,d

2.b,c,d

3.c,d,a

4.a,b,c

Solution- 4.a,b,c
Reason 1-Machine translation is the process of using computer programs to translate a text/speech
from one natural language to another relevant to context. It has the ability to translate in many
languages and also required for web content and web page translation

Q:57. A brief history of Machine Translation includes:

a. Rule based Machine Translation (RBMT)

b. Example based Machine Translation (EBMT)
c. Statistical Machine Translation (SMT)
d. Neural Machine Translation (NMT)

1.a,b,c

2.c,d,a

3.a,b,c,d

4.b,c,d

Solution- 3.a,b,c,d

Reason 1-RBMT (Rule based Machine Translation) was in 1950, EBMT (Example based Machine
Translation) was in 1980, SMT was in 1990, and NMT was in 2015.

Q:58. Which of the following includes major tasks of NLP?

1.Automatic Summarization

2.Natural language understanding

3.Natural language generation

4.All the above

Solution- 1. Automatic Summarization

Reason 1- By utilizing NLP, developers can organize and structure knowledge to perform tasks such
as automatic summarization, translation, named entity recognition, relationship
extraction, sentiment analysis, speech recognition, and topic segmentation.

Q:59. Google translator is the application of

1.Machine Translation

2.Text summarization

3.Information extraction

4.None of the above

Solution- 1.Machine Translation

Reason 1- Google translator is an application of Statistical and neural machine translation

Reason 2- Google has made Google translator application by using Statistical and neural machine
translation.

Q:60. Applications of NLP are

a. Chatbots
b. Voice assistants
c. Virtual assistant
d. None of the above

1.a,b,d

2.a,b,c

3.b,c,d

4.a,c,d

Solution- 2.a,b,c

Reason 1- All the three chatbot, virtual assistant and voice assistant are the application of NLP.

Q:61. Applications of Deep Learning are:

1.Self-driving cars

2.Fake news detection

3.Virtual Assistants

4.All the above

Solution- 4. All the above

Reason- Self-driving cars, Fake news detection and virtual assistants all are the applications of deep
learning including healthcare, fraud detection etc..

Q:62. The inputs for a single layer neural network are 1, 3, 2 and the weights of links connecting
input neurons to the output neuron are 2, 2, and 3 then the output will be (Identity activation
function is used in output neuron):

1.6

2.14

3.12

4.None of the above

Solution- 2. 14
Reason- by using this formula – Output = w1 * x1 + w2 * x2 + w3 * x3

Q:63. Which of the following is not a type of Artificial Neural Network?

1.Perceptron

2.Radial Basis Functions

3.Random Forest

4.Autoencoder

Solution- 3. Random Forest

Reason- Both the Random Forest and Neural Networks are different techniques that learn
differently but can be used in similar domains. Random Forest is a technique of Machine Learning
while Neural Networks are exclusive to Deep Learning.

Q:64. What is the limitation of deep learning?

1.Amount of data

2.Computational expensive

3.Data Labeling

4.All the above

Solution- 4. All the above

Reason- Amount of data, computational expenses and data labelling all three are the limitation of
deep learning.

Q:65. The number of nodes in the hidden layer is 8 and the output layer is 5. The maximum number
of connections from the hidden layer to the output layer are:

1.40

2.Less than 40

3.More than 40

4.It is an arbitrary value

Solution- 1. 40

Reason- it is a fully connected direct graph, the number of connections are multiple of nodes in
hidden layer and output layer.

Q:66. Recurrent Neural Networks (RNN) are used for

1.Businesses Help securities traders to generate analytic reports

2.Detecting fraudulent credit-card transaction

3.Providing a caption for images

4.All of the above

Solution- 4. All of the above

Reason- All of the above options are true. This is because RNN is used to Help securities traders to
generate analytic reports, Detecting fraudulent credit-card transaction and to provide a caption for
images.

Q:67. Types of RNN are:

1.LSTM

2.Boltzman machine

3.Hopfield network

4.a and b

Solution- 4. a and b

Reason- A Boltzmann machine (also called stochastic Hopfield network with hidden units or
Sherrington–Kirkpatrick model with external field or stochastic Ising-Lenz-Little model) is a type of
stochastic recurrent neural network. LSTM networks are a type of RNN that uses special units in
addition to standard units.

Q:68. What is perceptron?

1.a single layer feed-forward neural network with pre-processing

2.an auto-associative neural network

3.a double layer auto-associative neural network

4.a neural network that contains feedb

Solution- 1. a single layer feed-forward neural network with pre-processing

Reason- It is the simplest type of feedforward neural network , a feedforward neural network with
no hidden units. Thus, a perceptron has only an input layer and an output layer.

Q:69. Which of the following architecture has feedback connections?

1.Recurrent Neural network

2.Convolutional Neural Network

3.Restricted Boltzmann Machine

4.None of these
Solution- 2.Convolutional Neural Network

Reason- CNN is a feed forward neural network that is generally used for Image recognition and
object classification.

Q:70. Bidirectional RNN:

1.Trained to predict both the positive and negative directions of time simultaneously.

2.Applications are speech recognition, handwritten recognition etc.

3.After forward and backward passes are done, the weights are updated

4.All the above

Solution- 4. All the above

Reason- All the above options are true.

Q:71. The incorrect statement for a Convolutional Neural Network are:

1.The height and width of the filter in CNN must be less than the size of input

2.The Pooling layer progressively increases the spatial size of the representation

3.It uses both linear and non-linear activation functions

4.The last few layers are fully connected layers and computation on these layers are very time
consuming

Solution- 1.The height and width of the filter in CNN must be less than the size of input

Reason- This is because the height and width of the filter in CNN must not be less than the size of
input.

Q:72. A Convolutional Neural Network is able to successfully capture the Spatial and Temporal
dependencies:

1.True

2.False

Solution- 1.True

Reason- Yes this true, because a ConvNet is able to successfully capture the Spatial and Temporal
dependencies in an image through the application of relevant filters.

Q:73. Different types of normalization in Deep Neural Networks are

a. Output
b. Batch
c. Group
d. Instance
1.a,b,c

2.b,c,d

3.d,a,b

4.d,a,c

Solution- 2.b,c,d

Reason- Different types of normalization in Deep Neural Networks are batch, group, instance, layer,
weight etc..

Q:74. Applications of CNNs are:

a. Recommender systems
b. AlexNet
c. Natural Language Processing
d. Pooling

1.a,b

2.b,d

3.a,c

4.a,d

Solution- 4.a,d

Reason- This is because recommender system is an application of CNN and Pooling layers are used
to reduce the dimensions of the feature maps in CNN.

Q:75. Which of the following statements are correct for GAN?

a. GANs are useful for unsupervised learning, supervised learning, semi-supervised learning, and
reinforcement learning
b. Generative model technique learns to generate the data with the same statistics of training data
c. At each iteration the goal of generator is to minimize the classification error and the goal of
discriminator is to maximize the classification error.
d. The discriminator could tell the difference between images of a cat and a dog and generative
model could generate new images of animals that look like real animals.

1.a,b,c

2.a,b,d

3.a,c,d

4.b,c,d

Solution- 2.a,b,d
Reason- All the three a, b and d options are true.

Q:76. A generative model:

a. Captures the joint probability p(X,Y)

b. Captures the conditional probability p(Y|X)
c. Includes the distribution of data itself
d. Cannot predict the next word in sequence

1.a,b

2.a,c

3.a,d

4.b,c

Solution- 2. a, c

Reason- A generative model includes the distribution of the data itself, and tells you how likely a
given example is. For example, models that predict the next word in a sequence are
typically generative models (usually much simpler than GANs) because they can assign a probability
to a sequence of words. and it. It also captures the joint probability p(X,Y)

Q:77. The discriminative model:

a. Draw boundaries in the data space as it tells the difference between handwritten 0s and 1s.
b. Captures the joint probability p(X,Y)
c. Captures the conditional probability p(Y|X)
d. Learns to distinguish the generator’s fake data from real data

1.a,b,c

2.a,b,d

3.a,c,d

4.b,c,d

Solution- 3. a,c,d

Reason- The discriminative model tries to tell the difference between handwritten 0’s and
1’s by drawing a line in the data space. It captures the conditional probability p(Y|X) and Learns to
distinguish the generator’s fake data from real data

Q:78. Choose the incorrect statements from the following

1.The discriminator uses the real data as negative examples during training
2.The discriminator uses the fake data as negative examples during training

3.The portion of the GAN that trains the generator model includes random input

4.None of the above

Solution- 1.The discriminator uses the real data as negative examples during training

Reason- The generated instances become negative training examples for the discriminator. The
discriminator learns to distinguish the generator’s fake data from real data.

Q:79. Choose the correct statements from the following

a. Most universal approximation theorems can be parsed into two classes. The first quantifies the
approximation capabilities of neural networks with an arbitrary number of artificial neurons and the
second quantifies an arbitrary number of hidden layers
b. A neural network can represent any function provided it has sufficient capacity.
c. Not all architectures can represent any function
d. None of the above

1.a,b,c

2.b,c,d

3.d,a,b

4.d,a,c

Solution- 1.a,b,c

Reason- All the above options a,b and c are true. As Most of the universal approximation theorems
can be parsed into two class. The first class quantifies the approximation capabilitiesof neural
networks with an arbitrary number of artificial neurons and the second quantifies an arbitrary
numberof hidden layers

Q:80. Interesting applications of Generative Adversarial Networks (GANs) are:

1.Photo Inpainting

2.Culinary arts (as making a pizza)

3.Face aging

4.All the above

Solution- 4. All the above

Reason- Photo Inpainting, Culinary arts (as making a pizza), Face aging all are the applications of
Generative Adversarial Networks (GANs)

Q:81. Applications of face recognition system are

a. Security system
b. Attendance system
c. Smartphone’s unlocking system
d. Video surveillance system

1.a,b,c,d

2.a,b,c

3.b,c,d

4.d,a,b

Solution- 1.a,b,c,d

Reason- Attendance system, video surveillance system, Smartphone’s unlocking system and Security
system all 4 are the application of face recognition system.

Q:82. Image recognition steps are

a. Extract pixel features from an image

b. Prepare labeled images for training data
c. Train the model to be able to categorize images
d. None of the above

1.a,b,c

2.b,c,d

2.d,a,b

2.d,a,c

Solution- 1. a,b,c

Reason- All the above options a, b and c are true.

Q:83. Applications of image recognition system are

a. Driverless car technology
b. Document clustering
c. Gaming
d. All the above

1.a,b

2.b,c

3.a,c

4.c,d

Solution- 3. a, c
Reason- Developers of self-driving cars use vast amounts of data from image recognition systems,
along with machine learning and neural networks, to build systems that can drive autonomously.
The gaming arena has started using image recognition technology coupled with augmented reality
to their advantage, as it helps to provide gamers with a realistic experience.

Q:84. Object detection

a. It is a computer vision technique that allows us to identify and locate objects in an image or video
b. Object detection allows us to at once classify the types of things found while also locating
instances of them within the image
c. Draws a box around each object whereas image recognition assigns a label to an image
d. It has a capability which enables a program to process human speech into a written format

3.a,c,d

1.a,b,c

2.b,d,a

4.b,c,d

Solution- 1.a,b,c

Reason- Locate the presence of objects with a bounding box and types or classes of the located
objects in an image. It also allows us to identify and locate objects in an image or video. It allows us
to at once classify the types of things found while also locating instances of them within the image.

Q:85. Applications of object detection are

a. Video surveillance

b. Crowd counting

c. Spam/Non-spam classification

d. Self-driving cars

1.a,b,c

2.a,b,d

3.a,c,d

4.b,c,d

Solution- 2. a,b,d

Reason- Object Detection algorithms detect various types of objects for video surveillance
applications. The Nanonets API allows you to build Object Detection models with ease in crowd
counting. It is also used in self driving cars.

Q:86. Methods for object detection are

1.Viola–Jones method
2.Deep Learning approaches

3.Both (a) and (b)

4.None of the above

Solution- 3.Both (a) and (b)

Reason- The Viola–Jones object detection framework is an object detection framework which was
proposed in 2001 by Paul Viola and Michael Jones. Deep learning approach, is widely used one, for
detecting objects in images

Q:87. In which step of the processing, assigning a label (e.g., “dog”) to an object based on its
descriptors is done?

1.Object recognition

2.Morphological processing

3.Segmentation

4.Representation

Solution- 1. Object recognition

Reason- Object recognition step of the processing, assigning a label (e.g., “dog”) to an object based
on its descriptors is done

Q:88. What is the first and foremost step in Image Processing?

1.Morphological processing

2. Compression

3.Image acquisition

4.Image enhancement

Solution- 3. Image acquisition

Reason- Image acquisition is the first and foremost step in Image Processing.

Q:89. Image recognition tools are:

a. Clarifai

b. IBM Watson Visual Recognition

c. Amazon Lex

d. Scikit-image

1.a,b,c
2.b,c,d

3.a,b,d

4.d,a,c

Solution- 3.a,b,d

Reason- scikit-image is a collection of algorithms for image processing. IBM Watson Visual
Recognition is a tool that allow users to automatically identify subjects and objects contained within
the image and organize and classify these images into logical categories. By leveraging the power
of Clarifai’s image classification tools you can train your apps to recognize just about anything.

Q:90. Choose an incorrect statement

1.Image detection involves predicting the class of one object in an image.

2.Object localization refers to identifying the location of one or more objects in an image and
drawing abounding box around their extent

3.Object detection combines these two tasks and localizes and classifies one or more objects in an
image

4.Detection is the process of identification and classification is the categorization of the object based
on a previously defined classes or types

Solution- 2. Object localization refers to identifying the location of one or more objects in an
image and drawing abounding box around their extent

Reason- All the above are correct except the second one.

Q:91. Choose the incorrect statement:

1.Speech recognition is a way of encoding and decoding analog signals

2.Differently abled people can use speech recognition system.

3.The first speech recognition systems were focused on numbers not words

4.None of the above

Solution- 4. None of the above

Reason- This is because speech recognition is a way of encoding and decoding analog signals, and it
can be used for Differently abled people. The first speech recognition systems were focused on
numbers, not words. In 1952, Bell Laboratories designed the “Audrey” system which could
recognize a single voice speaking digits aloud.

Q:92. Voice search engines are:

a. Google Assistant
b. Siri
c. Alexa
d. Cortana
1.a,b,c,d

2.b,c,d

2.d,a,b

2.d,a,c

Solution- 1. a,b,c,d

Reason- All the four are voice search engines given in the options

Q:93. Types of speech recognition systems

1.Speaker independent

2.Speaker dependent

3.Speaker adaptive

4.All the above

Solution- 4. All the above

Reason- if speaker-dependent data are available, the system could be adapted to the
specific speaker such that the error rate could be significantly reduced. Speaker–
dependent software is commonly used for dictation software, while speaker–
independent software is more commonly found in telephone applications.

Q:94. Steps of speech recognition system include

a. Analog to digital conversion

b. Wavelets and multiresolution processing
c. Use of acoustic and language model
d. Segmentation

1.a,b

2.b,c

3.c,a

4.a,d

Solution- 4. a,d

Reason- First, the speech is analyzed to determine the likely locations of phoneme boundaries;
second, the segments that result from this analysis are classified based on features taken from
throughout the segment

Q:95. Applications of speech recognition systems are

a. Home automation
b. Voice dialing
c. Color processing
d. Translation

1.a,b,c

2.a,b,d

3.a,c,d

4.b,c,d

Solution- 2.a,b,d

Reason- Speech recognition applications include voice user interfaces such as voice dialing (e.g.
“call home”), call routing (e.g. “I would like to make a collect call”) and translation.

Q:96. In computer vision

a. The tasks include methods for acquiring, processing, analyzing and understanding digital images
b. Most computer vision systems rely on image sensors, which detect electromagnetic radiation,
which is typically in the form of either visible or infra-red light
c. The working of visual cortex of a dog has introduced the concept of edge detection.
d. All the above

1.a,b

2.b,c

3.c,d

4.a,d

Solution- 1.a,b

Reason- The option a and b are true because computer vision tasks include methods for acquiring,
processing, analyzing and understanding digital images and most computer vision systems rely on
image sensors, which detect electromagnetic radiation, which is typically in the form of either visible
or infra-red light

Q:97. How computer vision works

a. Acquiring an image
b. Processing the image
c. Analyzing the image
d. Understanding the image

1.a,b,c,d
2.a,b,c

3.d,a,b

4.d,a,c

Solution- 1. a,b,c,d

Reason- Computer vision tasks include methods for acquiring, processing, analyzing and
understanding digital images, and extraction of high-dimensional data from the real world in order
to produce numerical or symbolic information

Q:98. Potential application areas of robots

a. Military robots
b. Drones
c. Medical robot
d. None of the above

1.a,b,c

2.b,c,d

3.a,b,d

4.d,a,c

Solution- 1.a,b,c

Reason- All of the above options are the potential applications of robots.

Q:99. Artificial Intelligence techniques can

a. Analyze large amount of data

b. Take complicated decision easily
c. Replace humans in near future from almost all type of jobs
d. Be used in Logistic and Supply chain management

1.a,b,c

2.b,c,d

3.a,b,d

4.d,a,c

Solution- 3.a,b,d

Reason- This is because experts are confident that artificial intelligence will operate hand in hand
with humans in the workplace, not take their jobs.
Q:100. How a robot works?

1.Uses pre-programmed instructions stored in CPU

2.It need special hardware with sensors and effectors

3.Robot sensors send feedback to controllers

4.All the above

Solution- 4.All the above

Reason- All the above options are true because robots uses some programs and it need special
hardware with sensors and effectors and these robot sensors send feedback to controllers.