IBM SkillsBuild

Artificial Intelligence Fundamentals

Course - 1
Introduction to Artificial Intelligence
1. Which of the following Narrow AI systems would you want to include when building the Broad AI system for a self-driving
self car?

a) Information from the global positioning system (GPS) b) Long-term weather patterns
c) International news d) Entertainment trends

2. Fill in the blank. __________ systems are used to predict global weather, trace pandemics, and help businesses predict future
f trends?

a) Targeted AI b) Narrow AI
c) Broad AI d) General AI

3. Complete the sentence. ________________ performs tasks without human intervention and completes mundane and repetitive taskst f
humans, while ________________ allows humans to make final decisions after analysing data, reports, and other types of data.

a) AI, augmented intelligence b) Augmented intelligence, AI

c) Augmented intelligence, Narrow AI d) AI, Narrow AI

4. Complete the sentence. Artificial intelligence (AI) refers to the ability of a machine to ________________ and ________________.
_________

a) learn patterns, make predictions b) identify emotions, transfer data

c) create algorithms, generalize information d) secure data, create algorithms

5. What one word best describes what AI can do?

a) Think b) Decide c) Calculate d) Predict

 1. In which era did computers guide astronauts from Earth to the moon and advance to handle more than one kind of instruction?
instructi

a) The Era of AI b) The Era of Industrialization

c) The Era of Tabulation d) The Era of Programming

2. In which era did the advancements in fields ranging from cancer research to energy production show how its achievements impact peo
very day, and that it has exponentially increasing abilities?

The Era of Tabulation b) The Era of Programming c) The Era of AI d) The Era of Industrialization

3. In which era did machines help humans sort data into structures to reveal its secrets, that is, to reveal more insight than
t just simply cou
get a sum total?
a) The Era of AI b) The Era of Programming
c) The Era of Industrialization d) The Era of Tabulation

4. What were the main causes of the First Winter of AI?

elect the two that apply.
a) Lack of funding b) Intellectual property theft
c) Buggy program code and slow language execution d) High expectations

5. Which of the following is the correct order of the three eras of computing?
a) Tabulation, AI, Programming b) Tabulation, Programming, AI
c) Programming, Tabulation, AI d) AI, Programming, Tabulation
1. Which of the following is most likely to be unstructured data?

a) The breeds of entrants in a dog show b) Ingredients for baking bread

c) Prescriptions for patients in a large hospital d) Social media posts

2. If names and email addresses are organized using a spreadsheet software, like Google Sheets or Microsoft Excel, which of type of dat
is?
a) Semi-structured data b) True data c) Structured data d) Unstructured data

3. Shelby writes a survey that includes the following question: “Do you like chocolate (circle Yes or No), and why?”
Which of the following describes the data Shelby will collect?

a) Structured data b) Both structured data and unstructured data

c) Unstructured data d) Neither structured data nor unstructured data

4. Fill in the blank. Data that is not completely raw and contains elements such as tags and organizational metadata is known
kno as
_______________ .
a) semi-structured data b) unstructured data
c) structured data d) true data

5. Approximately how much of the world’s data today is unstructured?

a) 85% b) 90% c) 80% d) 40%
1. Fill in the blank. Machine learning can analyze ______________ data more quickly than a programmable computer can.

a) Supervised b) triangulated c) structured d) dark

2. Which of the following is an example of a probabilistic calculation?

a) A 78% chance of being wrong b) Wording that is probably correct

c) A strong feeling you are most likely correct d) An unreliable calculation

3. Complete the sentence. Machine learning is a type of AI with many advantages over programmable computers, including the distinguis
capability to _______________ and _______________.
a) use structured data, locate solutions b) search, define
c) predict, learn d) use databases, find solutions

4. Tomas is programming an app to help people find their way in his city. He programs every possible route from the university
univers auditorium
museum.
Which of the following is Tomas likely using?
a) Deep learning b) Heuristic computing c) Machine learning d) Programmable computing

5. Why does machine learning find answers in unstructured data more quickly than a programmable computer?
a) Machine learning can search every combination of factors very quickly.
b) Machine learning analyzes every data combination.
c) Machine learning uses the most sophisticated program code.
d) Machine learning assembles all the data into a table before calculating.
1. Fill in the blank. ______________ learning occurs when the algorithm is given unlabeled data.

a) Unsupervised b) Classical c) Reinforcement d) Supervised

2. Which method of learning does an AI system use when it uses the process of trial and error?

a) Supervised learning b) Classical conditioning

c) Reinforcement learning d) Unsupervised learning

3. YouTube and Netflix often recommend options to users.

What kind of learning is used to generate these recommendations?

a) Commentary learning b) Unsupervised learning

c) Supervised learning d) Reinforcement learning

4. You train Google Photos to identify people in photographs. As a result, Google Photos can recognize their faces. What kind
ki of learning

a) Reinforcement learning b) Transitory learning

c) Supervised learning d) Unsupervised learning

5. Fill in the blank. All supervised learning algorithms need ___________ data.

a) unlabeled b) unstructured c) labelled d) raw

 1. Which of the following represents the idea that AI will move into all industries, from finance, to education, to healthcare,
healthc increasing prod
nd enabling new opportunities?
a) Broad AI b) AI everywhere
c) Programmable AI d) Future AI

2. What do AI researchers call the type of systems likely to emerge by 2050 and beyond?

a) commentary learning b) Universal AI

c) Broad AI d) General AI

3. With the use of billions of sensors, what is the new idea that will help to improve the Earth’s safety, sustainability, and security?

a) Instrumented planet b) Narrow AI

c) Tabulated Earth d) Personalization

4. What type of AI do services, such as Netflix, use to recommend offerings to users?

a) Future AI b) Broad AI c) General AI d) Narrow AI

5. Tesla cars are all connected and share what they learn when autonomously driving. This is an example of which of the following
fol levels

a) Programmable AI b) General AI c) Future AI d) Broad AI

1. Which of the following describes how machine learning differs from classical systems in structuring unstructured data?

a) Machine learning constructs a predetermined structure of routes based on a database.

b) Machine learning relies on human judgment to structure unstructured data.
c) Machine learning offers probabilistic statements instead of binary decisions.
d) Machine learning relies on deterministic algorithms to process unstructured data.

2. Which level of AI refers to machines that can perform any intellectual task that a human can?
a) Narrow AI b) Broad AI c) General AI

3. Which of the following types of data can be analyzed more quickly by machine learning than by a programmable computer?

a) Systematic b) Structured c) Systematic d) Unstructured

4. Which of the following is a likely consequence of AI becoming more pervasive across industries?
a) Reliance on current machine interface technologies will strengthened
b) Enhanced productivity and the creation of new opportunities
c) Increased compartmentalization of AI implementation
d) Humans and machine interactions will become increasingly standardized

5. Which of the following statements most accurately describes semi-structured

semi data?
a) Data that lacks any built-in
in organization or structure, such as images and texts
b) Quantitative data processed and analyzed by conventional data tools
c) Information that can be organized in rows and columns, like in a spreadsheet
d) Data that combines features of both structured and unstructured data, using metadata for organization
 6. Which of the following is the level of AI that is more versatile, can handle a wider range of related tasks, and is focused
focu on integrating A
pecific business processes?

a) Broad AI b) General AI c) Narrow AI

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

a) Flight schedules b) Videos c) Social media posts d) Movie reviews

8. Which of the following types of data is categorized as qualitative data, and cannot be processed and analyzed by conventional
convent data too
methods?
a) Unstructured data b) Semi-structured data c) Data in a database d) Structured data

9. In which of the following eras of computing did an IBM system beat a chess champion, a robot drive itself down a 131-mile
131 desert trail, a
IBM Watson beat two Jeopardy! grand champions?

a) Era of Tabulation b) Era of AI c) Era of Programming d) The Era of Industria

10. Which of the following explains why an AI trained with machine learning can rapidly map the best route between two points
poin on a map?

a) It uses predefined programming instructions to navigate through traffic.

b) It performs millions of tiny calculations quickly, trying different routes through trial and error.
c) It relies on a complete database of all possible routes through the city.
d) It predicts traffic problems based on stored alternative routes for every possible situation.
11. Which of the following is a key attribute of an artificial intelligence system?
a) AI systems are self-aware. b) AI systems can judge.
c) AI systems have experience. d) AI systems make predictions.
12. Which of the following acts on behalf of people in the physical world in a way that complements human abilities?

a) Cognitive computing systems b) Expert systems

c) Machine learning algorithms d) Augmented intelligence

13. You are working on an AI system to help doctors treat patients. While describing your system to a group of doctors, you are asked the
following question, “How does machine learning enable decision-making
making in complex scenarios, such as medical treatment choices?”
a) By offering deterministic solutions based on fixed criteria
b) By providing probabilistic statements with confidence levels for different treatment options
c) By relying solely on AI systems to make treatment decisions without human input
d) By ignoring potential outcomes and personal experience in decision making
14. Which of the following methods of learning describes how an AI system learns using trial and error?
a) Supervised learning b) Reinforcement learning c) Unsupervised learning

15. You’re working for an employer who wants to learn from social media what customers say about her products and services. Which of
following is the best reason to support moving to an AI system?
a) Social media data lacks complexity.
b) Social media data is completely structured.
c) Social media data is completely unstructured.
d) Social media data contains a mix of structured and unstructured data.
 IBM SkillsBuild

Artificial Intelligence Fundamentals

Course - 2
Natural Language Processing and Computer Vision
Q 1. Which of the following functions can a cognitive AI system perform? Select all that apply.

a) Reason b) Intend c) Understand d) Interact e) Learn

Q 2. Fill in the blank. A system that helps machines understand cultural knowledge, grammatical structures, and various language
langu
nuances is known as _____________________.
a) natural language processing b) relational databases
c) translators d) computer vision

Q 3. Which of the following statements is correct about natural language processing (NLP)? Select all that apply.
a) IBM Project Debater was a major step forward in the field of natural language processing.
b) Human language is complex, so machines require natural language processing systems to understand human langua
c) IBM Project Debater was close, however not a successful project in the field of natural language processing.
d) In natural language processing, an AI system can speak human language perfectly once it builds a corpus.

Q 4. What term do AI researchers use to describe the collection of text and data they use to train a debater AI system?
a) Corpus b) Compilation c) Collation d) computer vision
5. What is the final step an AI debater system must take?

a) Organize the proof by deciding what evidence is the strongest, arrange it by themes, and adjust the arrangement for new evide
score points against the opponent.
b) Respond to the opponent by listening to the opponent's arguments to then deliver a convincing rebuttal that refutes th
opponent and further proves the case.
c) Build the final position by creating a final speech that details the position on the debate topic and gives a compelling argument.
a
d) Ingest a collection of learned material to then structure the content to evaluate concepts.
Q 1. What type of problem does the following sentence create for an AI system? Is your refrigerator running?
a) Classification problem b) Token movement problem
c) Collation d) Sentiment problem
Q 2. How does an AI system that is working with human language handle the classification problem?
a) It consults the internet for interpretation. b) It segments the associated tokens.
c) It learns from many instances. d) It creates a reference database.

Q 3. What is sentiment analysis?

a) A method to determine the emotion when AI reads data, such as a customer service chat
b) A method to identify data as angry, cheerful, or scared
c) A method to measure the strength of an emotion by assessing data as positive, negative, or neutral
d) A method to assess data into distinct human emotion types

Q 4. You’re consulting for a store that wants to know how customers feel about a new product line. You propose an AI system that
t can
study how people feel about these products based on their comments in social media.
Which tokens are most important for the AI system’s NLP to analyze?
a) Concepts and classification b) Entities and relationships
c) Emotions and sentiment d) Loyalty and sentiment
Q 5. Your friend says, “I really love anchovies and flowers.” How many relationships between entities can you find in this statement?
st
a) Two b) Four c) One d) Three
Q 1. For a chatbot,, what is an entity considered within the context of a question?
a) A process b) An adjective c) A node d) A noun
Q 2. Which of the following statements describes how a chatbot processes a question?
a) The chatbot breaks inputs apart to identify intents and entities.
b) The chatbot combines entities to find how relationships divide at nodes.
c) The chatbot logs each question into a frontend database.
d) The chatbot creates frontend triggers.
Q 3. The majority of callers to a local hospital want to do one of the following three tasks: make an appointment, cancel an appointment,
and get the hospital’s address. Would this be an appropriate use for a chatbot and why?
a) Yes, because chatbots are designed for a small set of interaction types.
b) Yes, because chatbots were specifically invented to handle hospital inquiries.
c) No, because chatbots are designed for very large sets of interaction types.
d) No, because chatbots are intended for a wider range of responses.
Q 4. Fill in the blank. A dialog is represented by a ____________ that illustrates how a chatbot will respond after a human asks a questio
a) detailed bulleted list b) flowchart c) movie script d) three-dimensional ta

Q 5. Fill in the blank. A chatbot runs ________________ to understand the human’s intentions of their typed input.

a) transistors b) translators
c) natural language processing (NLP) d) language processing pods (LPP)
Q 1. Which of the following would be a good use of computer vision in a self-driving
self car?
a) Detecting trunk capacity b) Detecting the speed
c) Detecting tire pressure d) Detecting a stop sign
Q 2. If a doctor asks you to suggest an AI capability that can help analyze X-rays,
X which computer vision technology would you suggest
a) Generative adversarial network (GAN) b) Heads up guidance system (HGS)
c) Convolutional neural network (CNN) d) Natural language processing (NLP)
Q 3. If you were a consultant, which of the following applications would you recommend for using computer vision?

a) Creating realistic portraits of historical figures b) Designing cable structures for a suspension bridge
c) Inspecting robots at work on an assembly line d) Writing advertising text for a movie studio

Q 4. A generator creates an image of a flower and the discriminator determines if the image is real or fake. This process is repeated unt
the generator is able to create a image of a flower good enough to fool the discriminator. This is an illustration of which of
o the follow
machine learning models?
a) Heads up guidance system (HGS) b) Convolutional neural network (CNN)
c) Generative adversarial network (GAN) d) Natural language processing (NLP)

Q 5. Which process makes facial recognition possible?

a) Convolutional neural network (CNN) b) Generative adversarial network (GAN)
c) Natural language processing (NLP) d) Heads up guidance system (HGS)
Q 1. Which of the following explains how detecting emotions can help an AI system in user interactions?
a) It allows the AI to predict the user’s future actions and purchasing decisions.
b)) It enables the AI to detect when a customer is frustrated or angry and escalate the issue to a human agent
c) It allows the AI to learn the user’s preferred language and communication style.
d) It helps the AI identify the user’s age and demographic information.

Q 2. Which of the following processes is used by AI to analyze only small parts of an image at a time, making it possible for visual
recognition systems to identify parts of an image?
a) Convolutional neural network (CNN) b) Natural language processing (NLP)
c) Generative adversarial network (GAN) d) Vocal recognition network (VRN)

Q 3. Complete the sentence. _______________ can be more difficult for an AI system than identifying tokens because so much of it dep
on the context of the sentence. For example, compare “I went to the docks to ship my box” to “I went to the station to ship my box”.
a) Classification b) Sentiment analysis c) Entity identification d) Sentence segmentation

Q 4. Which of the following grammatical terms best describes an entity in relation to chatbots?
a) Noun b) Verb c) Preposition d) Adjective
Q 5. Natural language processing is the field of computer science concerned with which of the following?
a) Translating between different programming languages
b) Composing music based on other creative works
c) Understanding and manipulating spoken and written language
d) Building robots that can move and interact with the physical world
Q 6. Imagine you want to create a new AI application. Which of the following would be the most suitable use of NLP for the application?
a) An AI-powered
powered financial fraud detection systems that analyzes transaction patterns
b) An AI-powered customer service chatbot that can handle complex inquiries and resolve customer issues
c) An AI-powered
powered music composer that generates original music pieces based on other compositions
d) An AI-powered
powered stock market prediction tool that analyzes stock prices
Q 7. You are reading reviews of the latest blockbuster movie. One review says, “This movie is fantastic!” Another review says,
says “Worst movi
ver!” Which of the following best describes the process that an AI system would use to classify the reviews?
a) Keyword analysis b) Topic modeling c) Topic modeling d) Sentiment analysis
Q 8. Which of the following is the messaging channel of a chatbot that interacts with the user who is asking the question?
a) Backend b) API gateway c) Natural language processing engine d) Frontend

Q 9. Which of the following terms is used to describe the reason why a user sends a message in a chatbot?

a) Entity b) Sentiment c) Intent d) Dialog

Q 10. Which of the following describes how chatbots differ from more general AI systems in their capabilities?
a) Chatbots are more intelligent and adaptable than general AI systems.
b) Chatbots can access and process more information than general AI systems.
c) Chatbots are focused on answering specific questions in a limited domain.
d) Chatbots rely on real-time
time data input while general AI systems use pretrained data sets.

Q 11. Which of the following is the best use of computer vision in a self-driving
self car?
a) Detecting a stop sign b) Determining the car’s route
c) Detecting tire pressure d) Detecting speed

12. Which of the following is a convolutional neural network (CNN) used for when an AI system processes images to derive meaning from
hem?

a) Generating random patterns in images

b) Analyzing entire images at once
c) Supervising the training process and updating the model’s parameters
d) Identifying patterns and features in images
13. Imagine you are the CEO of a large retail company. You are considering implementing chatbots on your website to interact with
customers. Which of the following situations would be the most appropriate use of a chatbot?
a) Handling situations where customers are unsure of what kind of support they need
b) Negotiating personalized discounts and deals with customers
c) Handling situations with customers that require empathy and judgment
d) Answering frequently asked questions (FAQs) about store policy, product availability, and return procedures

14. You are working on a project that has two convolutional neural networks (CNNs) working against each other. One CNN (the
generator) creates new data, while the other CNN (the discriminator) tries to identify if the data is real or generated. This competitive
process helps the generator create increasingly realistic outputs. Which of the following best describes what you are working on?
on
a) Variational Autoencoder (VAE) Transformer
b) Convolutional neural network (CNN)
c) Natural language processing (NLP) system
d) Generative adversarial network (GAN)

15. When handling unstructured information, computers begin with one sentence at a time using sentence segmentation. Computers
Comput
then break the information into smaller chunks that can be individually classified and sorted into a structure for natural language
la
processing to work with them. Which of the following is the term used to refer to those smaller chunks?
a) Tokens b) Entities c) Sentiments d) Intents
 IBM SkillsBuild

Artificial Intelligence Fundamentals

Course - 3
Machine learning and deep learning
Q 1. Which of the following statements is correct?
a) Reinforcement learning is classified by humans.
b) Unsupervised learning requires human input except on the raw data.
c) Supervised learning requires humans to provide labelled data.
d) Supervised learning requires no human input. except for the raw data itself.

Q 2. Aditi is training an AI system in which it is penalized for answers that are largely wrong and rewarded for answers that are largely
largel
correct. What type of machine learning is she using?
a) Supervised learning b) Structured learning
c) Reinforcement learning d) Unsupervised learning
Q 3. Pedro is working with a chess-playing
playing AI system. He rewards it when it wins and penalizes it when it loses. What kind of learning
l is
the system experiencing?
a) Structured learning b) Unsupervised learning
c) Supervised learning d) Reinforcement learning

Q 4. Katie has a huge amount of warehouse data to sort but doesn’t know how to categorize it. Which method of machine learning
learni
would help her the most?
a) Structured learning b) Unsupervised learning
c) Supervised learning d) Classical learning
5. Maria is involved in training the AI for her company. She separates weather data in a database into different categories.
categories What type
of learning is she preparing?
a) Supervised learning b) Unsupervised learning
c) Reinforcement learning d) Deep learning
Q 1. You’ve been asked to create a machine learning service that helps people choose what concert to attend on a particular date
d
based on the type of music they prefer, who is singing, and where the event is taking place. What type of algorithm should you
y use?

a) Logistic regression b) Decision tree c) Supervised learning d) Reinforcement learning

Q 2. Which of the following is an advantage of classical machine learning compared to deep learning?
a) It’s obsolete. b) It’s obsolete. c) It’s easier to interpret. d) It’s more expensive to operate.

Q 3. Which of the following is an important difference between linear and logistic regressions?
a) Only linear regressions have outputs shaped like an S-curve.
curve.
b) Only logistic regressions have outputs between 0 and 1.
c) Only logistic regressions follow branching logic paths to an answer.
d) Only linear regressions have outputs that cannot be resolved.

Q 4. You want to know how many hours to study to pass an exam. The “hours of studying” data forms a sigmoid function, or an S-
S
shaped curve, between “fail” and “pass”. Which learning algorithm could you use?
a) Logistic regression b) Decision tree
c) Stochastic learning d) Unsupervised learning
Q 5. What type of algorithm would help you find out what you might earn for a salary if you have different levels of college degrees in
computer science?

a) Neural network b) Linear regression

c) Supervised learning d) Logistic tree
Q 1. How can you demonstrate the concept of machine learning?

a) Determine the farthest distance from a target that an archer can reliably hit bulls eyes by shooting arrows repeatedly wh
walking closer and closer to the target.
b) Determine the best way to book airline connections by running through a decision tree of possible flights, destinations, and
a
departure dates and times.
c) Every day for two weeks, ask a librarian what books to read and keep a record of each recommendation.
d) Determine how hot a cup of coffee should be by consulting a list of preferred brewing temperatures

Q 2. After calculating values, what do neural networks often assign to those values that impact the final result?

a) Weight b) Cosine c) Angles d) Activation

3. You are designing a perception. Where will you locate the algorithms that run on the signal and then next passes the results
res to the
output layer?

a) In the nucleus b) In the neurotransmitter

c) In the input layer d) In the hidden layers
Q 4. Fill in the blank. In deep learning, many forms of _______________ devices make up the deep learning ecosystem.

a) deep capillary network b) deep neural network

c) deep ID net d) deep eco Net
Q 5. What kind of brain cells inspired the creation of neural networks?
a) Neuro bots b) Neurons c) Neuroses d) Neuralgias
Q 1. Which of the following is a limitation of generative AI? Select all that apply.
a) Royalty issues b) Lack of originality in content
c) Bias in the generated content d) Accuracy of content

Q 2. True or false? Roberto works as a coach on a professional sports team. One of his players is returning to workouts with the team
after being out 6 months due to an injury. Roberto could use generative AI to create a customized workout plan tailored to the
th returning
athlete's current fitness level.
a) FALSE b) TRUE
Q 3. Fill in the blank. A distinguishing feature of generative AI, as opposed to other types of AI, is its ability to _______________.
_________
a) use text as an input prompt b) recognize patterns
c) create new content d) return output related to the input
Q 4. Which of the following models does generative AI use to produce content? Select all that apply.
a) Distributive anchored categories (DACs) b) Autoregressive models
c) Generative adversarial networks (GANs) d) Variational autoencoders (VAEs)

Q 5. Which of the following is an example of generative AI?

a) Language translation b) Classifying data

c) Image recognition d) Composing music
Q 1. Fill in the blank. _______________ is a future form of artificial intelligence that will support a wide range of industries,
industr like
healthcare, science, and agriculture.

a) Broad AI b) AI everywhere c) AI targets d) AI future web

Q 2. Which one of the following is a new area of research that will extend the abilities of AI systems?

a) Isometric discovery b) Reinforcement learning

c) Academic enrollment d) Quantum computing
Q 3. Fill in the blank. In the future, _______________ is a capability that will enable AI systems to tailor recommendations and response
to you.
a) specialization b) personalization c) tailorization d) customization

Q 4. Which concept describes future AI’s ability to improve Earth’s safety, sustainability, and security?

a) GeoThrive planet b) Instrumented planet

c) Sustainable planet d) Earth Web planet

Q 5. What do AI researchers call the type of systems likely to emerge by 2050 and beyond?

a) Universal AI b) AI for Enterprise c) Broad AI d) General AI

Q 1. Henri is training an AI system. He has loaded the rules of the Chinese game of Go into the system. As the AI system plays
play its first fe
games, winning some and losing some, it gradually improves. What type of learning is Henri using to train the AI system?
a) Unsupervised learning b) Supervised learning c) Reinforcement learning
Q 2. Which of the following encompasses the use of neural networks with multiple hidden layers to analyze complex data and make
ma
predictions?

a) Deep learning b) Statistical methods for pattern recognition

c) Classical algorithms for data analysis d) Machine learning with single-layer neural networks

Q 3. Which of the following describes how information flows through the nodes of a perceptron in a neural network?
a) Information flows from the output layer to the hidden layers, where algorithms are applied, and then to the input layer.
b) Information flows from the input layer to the hidden layers, where algorithms are applied, and then to the output laye
c) Information flows directly from the input layer to the output layer, bypassing the hidden layers.
d) Information flows randomly between all layers of the perceptron.
perceptron
Q 4. Which of the following is an advantage of classical machine learning compared to deep learning?

a) Classical machine learning is easier to interpret.

b) Classical machine learning works well with unstructured data.
c) Classical machine learning is more expensive to operate, but gives better results.
d) Classical machine learning outperforms deep learning ecosystems.
Q 5. What is the primary source of inspiration for the design of neural networks?
a) Neural networks are inspired by the complex behaviours of animals in their natural environments.
b) Neural networks derive their design principles from mechanical engineering principles and systems.
c) Neural networks are primarily inspired by the workings of the human brain and its interconnected neurons
d) Neural networks draw inspiration solely from conventional computer architectures and algorithms.

Q 6. Complete the sentence. Deep learning can be differentiated from classical machine learning in that deep learning exclusively
exclusi uses
_______________.
a) Analyzing the correlation between income and education level.
b) Predicting the likelihood of customers purchasing a product based on their browsing history.
c) Determining the relationship between temperature and ice cream sales..
d) Predicting the grades of students based on the number of extracurricular activities they participate in.

Q 7. You’re writing a science fiction story in which the main character is travelling from Earth to the Andromeda galaxy. The main characte
ses the ship’s AI system to think just like a human to keep the ship safe throughout the voyage. Which of the following types
type of AI best
escribes the ship’s AI system?

a) Broad AI b) General AI c) Narrow AI

Q 8. In a bustling city, Emma, a real estate agent, is tasked with predicting housing prices for prospective buyers. She collects
coll data on
various factors such as square footage, number of bedrooms, neighbourhood, and distance to amenities. Emma wants to understand
understan
how changes in these variables impact the selling price of homes. Which of the following methods would Emma most likely use to
t
analyze the relationship between housing prices and factors such as square footage, number of bedrooms, and neighbourhood?
a) Linear regression b) Analytical regression c) Logistic algorithm d) Decision tree

Q 9. In a neural network, which of the following layers is where the activation algorithm is “fired”?

a) Axon layer b) Hidden layer c) Input layer d) Output layer

Q 10. Which of the following is a distinguishing feature of generative AI?

a) Generative AI relies solely on human input to generate content and does not have the capability to learn from data.
b) Generative AI generates new and original content based on a given prompt or data set.
c) Generative AI uses algorithms to identify patterns and relationships in existing data, but it cannot create entirely new content.
c
d) Generative AI predicts and classifies data such as discriminative AI models.
Q 11. Which of the following best describes how machine learning’s trial-and-error
trial learning process works?
a) Machine learning with single-layer neural networks
b) Machine learning learns by memorizing all possible outcomes and selecting the most common one as the correct answer.
c) Machine learning adjusts itself by constantly testing new data against its corpus, making modifications based on
established patterns.
d) Machine learning relies solely on preset algorithms to generate results without any adjustment or modification.

Q 12. Deep learning models are preferred over classical machine learning models in situations where which of the following occurs?
oc
a) The task involves simple classification problems.
b) The problem requires understanding of temporal dynamics such as in time series analysis.
c) The data is already in a structured format such as a CSV file or table.
d) The data set is very large and consists of unstructured data, such as images or text.

Q 13. Which of the following is revolutionizing industries by improving decision making, personalizing customer experiences, and
enhancing operational efficiency?
a) Generative AI b) Generative adversarial network (GANs)
c) Variational autoencoders (VAEs) d) Autoregressive models
Q 14. Complete the sentence. Deep learning can be differentiated from classical machine learning in that deep learning exclusively
exclus uses
_______________.
a) linear regression b) decision trees
c) artificial neural networks d) reinforcement learning

Q 15. As part of research team, you are tagging images of animals in a database as either “dogs” or “not dogs”. What kind of
o machine
learning are you preparing for?
a) Reinforcement learning b) Supervised learning c) Unsupervised learning
 IBM SkillsBuild

Artificial Intelligence Fundamentals

Course - 4
Run AI models with IBM Watson Studio
Q 1. You’re the project manager of an AI system designed to diagnose problems with fertilization in a large community of farmers.
farm You’v
hired agricultural experts; experts in the Python, Java, and Lisp programming languages; and data professionals who specialize
specializ in SQL
database queries. Meanwhile different farmers are submitting data on different crops but not being data experts, they are using
usi different
measuring systems at different times of year. Worst of all, they all work in different buildings, using different computer systems
sy that don’t
easily communicate with each other. Nothing is getting done.
Which capability of IBM Watson Studio would most help your AI development?

a) Easy visualization b) Efficient workflow

c) Collaborative environment d) Jupyter Notebooks

Q 2. Your client has requested that the AI system you’re working on should include a screen that displays several different graphs
g at one
time. What type of display is the client requesting?

a) A chronography b) An output collection

c) A dashboard d) An IDE

Q 3. What does IDE stand for?

a) Integrated development environment b) Internal development engagement

c) Interference deflection environment d) conventional computer programs can alter their own cod
Q 4. Fill in the missing word: A machine learning _______________ is a block of programming code.
a) functions

Q 5. Which of the following is a difference between conventional computer programs and machine learning models?

a) Conventional computer programs perform calculations.

b) A machine learning model delivers the same output every time.
c) A machine learning model can reprogram itself.
d) Conventional computer programs can alter their own code.
Q 1. When you selected the CSV file in All assets, what did you see?
a) A preview of the data to used when training your model b) A proposed name for your project
c) A set of provision options for Watson Machine Learning d) A list of possible storage location

Q 2. After you provisioned the database (Cloud Object Storage) what did you select to access AI/Machine Learning?

a) Resource list b) Code Engine c) Catalog d) Cloud Foundry

Q 3. Which of the following services did you provision first for your project?

a) Database storage b) Two competing resource names

c) All assets d) AI / Machine Learning

Q 4. After you finished setting up your IBM Watson Studio service, which of the following did you do next?

a) Set up the holdout and split b) Created a set of algorithms

c) Created a new project d) Downloaded historic data about loan defau

Q 5. Which of the following did you add your data set as in IBM Watson Studio?

a) Artifact b) Asset c) Project d) Resource

1. Why did Watson test four different algorithms for this AI model?

a) To identify those algorithms that triggered the confusion matrix

b) To remove the least effective model from the data set
c) To compare four different versions of the data set for accuracy
d) To determine which set of algorithms predicted defaults most effectively

2. Why did Watson Auto AI select Binary Classification as your Prediction Type?

a) Because predicting risk or no risk has only two options

b) Because you were dividing your training data into two portions
c) Because you really only needed two models to make your point
d) Because you only needed to manage Accuracy and Run time

3. In the simulation, you used training data and holdout data when setting up the details of the experiment. Ninety percent of the data wa
used to train the algorithms. Which of the following describes how the holdout data was used?

a) Training data b) Backup data c) Testing data d) Reserve data

4. What service did you associate with your experiment that helps with building a model?
a) AutoAI b) German_credit-data_training.csv
c) Sublimation matrix d) Machine Learning-Risk Fraud

5. Look at the confusion matrix generated by your AI model. Which of the following best describes the cell representing No Risk
Predicted and No Risk Observed?

a) True positive b) False positive c) True negative d) False negative

Q 1. Why would you save your model as a Jupyter Notebook (ipynb
ipynb) file? Select the two that apply .

a) To be able to edit the code and algorithms b) To secure the model against changes
c) To be able to use it with a different data set d) To encrypt the code for security purposes

Q 2. When you first save a notebook, it displays in editable form, but it is not _______________. Which option best completes
complete the
sentence?
a) trusted b) defined c) JavaScript d) part of the pipeline

Q 3. When your Pipeline notebook first displays, what form does it take?

a) View only b) JavaScript enabled c) Edit only d) Details undefined

Q 4. You’ve run several versions of an AI machine learning model for a client that wants to anticipate repair needs for a machine
mac on thei
assembly line. You’ve identified the version most likely to suit their needs. Now they're asking you to save it in its current
curren form witho
any changes. What does that tell you they plan to do with it?

a) Stop using it and delete it. b) Print it for publication in a paper-based journal.
c) Continue using it with additional data. d) Edit it in a Jupyter notebook using Python.

Q 5. Your AI team has developed several models to visually spot defects in building construction and trained them all on the same data
Now they’re about to generate a confusion matrix to select the best one. What steps will this involve?
a) Run them all on the identical batches of new data, then compare their performance.
b) Run three out of four model on new data, then compare them with the fourth as a control group.
c) Run them all and time how long it takes each one to create an accurate prediction.
d) Run them all on different data sets of data, average their outcomes, then compare the results.
 IBM SkillsBuild

Artificial Intelligence Fundamentals

Course - 5
AI Ethics
Q 1. Which of the following pillars of AI is exemplified by anyone being able to describe how an AI system makes a prediction or
recommendation?

a) Explainability b) Transparency c) Robustness d) Fairness e) Privacy

Q 2. An AI model is systematically giving an advantage or disadvantage to a certain group. This is an issue of which of the following?
f

a) Fairness b) Robustness c) Transparency d) Explainability e) Privacy

Q 3. _______________ is a multidisciplinary field that investigates how to optimize AI’s beneficial impact while reducing unintended
uni
or adverse outcomes.

a) AI Ethics

Q 4. A trustworthy AI model should allow key facts about that model to be seen very quickly, just like the nutritional label on a box of
food.
a) Privacy b) Transparency c) Explainability d) Explainability e) Robustness

Q 5. Which of the following pillars of AI ethics is most related to your AI model withstanding intentional and unintentional
interference?

a) Explainability b) Transparency c) Robustness d) Fairness e) Privacy

Q 1. When looking at a list of promotion candidates generated by an AI system, which of the following might be an indicator that
t
there is unwanted bias in the system? Select the two that apply.
a) A group receives a systematic advantage. b) All groups are represented proportionally.
c) All groups are represented equally. d) A group receives a systematic disadvantage.
Q 2. Complete the sentence. A _______________ group traditionally receives more favourable outcomes compared to a
_______________ group that traditionally receives less to no favourable outcomes.
a) unprivileged, privileged b) privileged, unprivileged c) protected, privileged d) privileged, unprotected

Q 3. According to the following graph, which of the following options is correct?

a) There is no privileged group. b) White employees belong to an unprivileged group.

c) White employees belong to a privileged group. d) Non-White employees belong to a privileged group.
Q 4. Which of the following is true about bias in AI systems?

a) Bias always enters AI systems through the actions of bad actors

b) All bias in AI systems results in unfair outcomes.
c) Bias in AI systems is always the result of race.
d) Unwanted bias is a systematic error in AI systems that may result in unfair outcomes.

Q 5. Fairness in AI aims to minimize which of the following?

a) Unwanted bias b) Unfavorable predictions

c) Protected attributes d) Favorable outcomes
Q 1. Which of the following is an example of poisoning in the AI context?
Select the three that apply.
a) Injecting malicious samples into the training data b) Corrupting test data sent to the deployed model
c) Corrupting test data sent to the deployed model d) Modifying AI model architecture, parameters, and logic

Q 2. In a poisoning attack, when does the adversarial attack occur?

a) Terminal phase b) Training phase c) Planning phase d) Post deployment phase

Q 3. Which of the following describes a robust AI system?

a) The AI system is secured on a computer behind locked doors and windows.
b) The AI system can effectively handle malicious attacks without causing unintentional harm.
c) The AI system is housed on a computer with lots of RAM and hard drive space
d) The AI system is very fast in processing data.

Q 4. Which of the following are examples of adversarial attacks on an AI system? Select the two that apply.

a) Terminal execution b) Evasion c) Denial of Service (DoS)

( attack d) Poisoning
Q 5. Which of the following describes how adversarial attacks are intentionally carried out on AI systems to accomplish a
malicious end goal?

a) Adversarial attacks attempt to clean data sets that the AI is using to train.
b) Adversarial attacks attempt to breach the secure facilities holding the computers the AI resides on.
c) Adversarial attacks attempt to exploit AI system vulnerabilities.
d) Adversarial attacks are only carried out by governments against governments.
Q 1. Which of the following options would be true for the AI system shown in the following graphic?

a) The AI system is transparent. b) The AI system is biased.

c) The AI system is very robust. d) The AI system has very little explainability.

Q 2. When developing explanations for the output of an AI model, it is important to remember that explanations must _______________
_________
depending on the user persona.

a) alter
3. Which of the following describes an AI system, that everyday people, who do not have any special training in AI, can understand
und
how and why the system came to a particular prediction or recommendation?

a) Transparent b) Interpretable c) Explainable d) Understandable

4. When doctors use AI to help make a medical diagnosis, it is important that the AI has which of the following?
elect the two that apply.

a) Explainability b) Examinability c) Indexability d) Interpretability

5. _______________ is the degree to which an observer can understand the cause of a decision.

a) Interpretability
Q 1. Sheldon is creating an AI recommendation system. Which of the following facts about the system should he share to ensure
transparency? Select the two that apply.
a) How decisions are made in the AI model b) Who has access to the data
c) How the AI’s neural network is structured d) What data is collected

Q 2. Complete the sentence. In order to be trustworthy, an AI system must be _______________ .

a) transparent b) flexible c) governable d) traceable

Q 3. _______________ is disclosing information related to the data used for building AI systems, design decisions made throughout
throug the
process, model creation, model evaluation, and model deployment.
a) Transparency

Q 4. Which of the following business roles is responsible for data? Select the three that apply

a) AI Ops Engineer b) Data Scientist

c) AI Model Lead d) Data Engineer

Q 5. Which of the following business roles contribute to design decisions? Select the two that apply.

a) Model Validator b) Business Owner

c) Data Engineer d) AI Model Lead
Q 1. Shirley was arrested for protesting when she in college. Which of the following would this be an example of?
a) Sensitive personal information (SPI) b) Personal information (PI)
c) Inferential personal information d) Harmful personal information

Q 2. Which of the following would not be considered personal information (PI)?

a) A person’s employer b) A person’s favourite colour

c) A person’s name d) A person’s postal code

Q 3. Ashbed wants to employ privacy controls before his model is deployed. Which of the following should he use? Select the two that
apply

a) Model anonymization b) Data minimization

c) Data anonymization d) Differential privacy

Q 4. Troy is trying to determine if Dean’s information was used in the training data set for a predictive AI system.
Which of the following describes the kind of attack Troy would use?
a) Personal information attack b) Group identity attack
c) Individual information attack d) Membership inference attack
Q 5. Annie is reviewing what data to collect for the AI model to process. She decides not to collect users’ mailing addresses.
addresses Which contr
is Annie employing?

a) Data anonymization b)) Data minimization

c) Differential privacy d) Model anonymization
Q 1. Daniel is trying to get into nursing school. His resume is excellent, but he was turned down by all the schools he applied
appli to. When h
looked into the admissions process, he found that all the schools used the same AI model to process the applications. When Daniel
Da
looked at the data for the past 10 years, he found that in every year no less than 60% of applicants were female and the rest were male
He also found that every year at least 90% of the students accepted were female. According to this data, which of the following
follow seems
be true?
a) Females are a privileged group. b) Males are a privileged group.
c) Females are an unprivileged group. d) Females and males are both unprivileged groups.

Q 2. Maria is on a team creating an AI model to help recruit engineers. Maria wants to ensure that the data set does not include
incl specific
attributes such as race, age, sex at birth, and ethnicity. What type of attributes are these examples of?

a) Privileged attributes b) Confidential attributes

c) Protected attributes d) Private attributes

Q 3. Which of the following adds random noise during model training to reduce the impact of any single individual on the model’s
mode
outcomes and to give a guarantee that an individual in the training data set could not be identified?

a) Referential reduction b) Differential privacy

c) Data minimization d) Model anonymization
Q 4. Clara has been brought in to evaluate an AI system. Based on the following graphic, she finds that Model B is more interpretable
inter than
Model A. Which of the following best describes why Model B is more interpretable than Model A?

a) Model B relies on a complex neural network structure for recommendations

b) Model B employs a sophisticated algorithm for recommendations.
c) Model B’s flow is clear and easy to understand.
d) Model B generates recommendations without any rationale.
Q 5. When you asked her why there were so many romantic comedies in her Netflix recommendations, your grandmother explained that
t
Netflix looks at the movies she watches and the movies she likes. She said they use that information to look for other movies to
recommend. Since she watches a lot of romantic comedies, she gets a lot of them recommended to her. Which pillar of AI ethics is
Netflix adhering to in this example?
a) Transparency b) Fairness c) Privacy d) Robustness e) Explainabili
Q 6. Rutherford is trying to implement a privacy control to fortify the AI system that he is working on. He decides to limit the amount
f personal and sensitive data that is collected and takes steps to ensure that the data collected is only as granular as needed.
nee
Which of the following privacy controls is Rutherford using?

a) Data aggregation b) Differential privacy

c) Model anonymization d) Data minimization

Q 7. Bao is reviewing a diagnostic AI model when he notices something strange. The data set is larger than it should be. After reviewing
reviewi
he activity over the past weeks, he notices that extra items were placed in the data set that could have resulted in the AI model rendering
ncorrect diagnoses. However, the AI model was able to process the bogus data and still render accurate and correct diagnoses for each
ata item.
his is an example of which of the following?

a) Transparency b) Fairness c) Privacy d) Robustness e) Explainability

Q 8. Which of the following business roles is responsible for the deployment of AI models?

a) Data Engineer b) Business Owner

c) AI Ops Engineer d) AI Designer
Q 9. Rose is working as an auditor on a large AI project. In an effort to provide transparency, she is creating a framework to
t map key
details of the system to the people who can provide that information.
Which of the following teams is responsible for the algorithms used to train the AI?

a) Design team b) Data team c) Model team d) Deployment team

Q 10. Uma works for a large travel booking company. She is responsible for disclosing all information related to the data her company use
u
o build their vacation recommendation AI system.
Which of the following is Uma in charge of?

a) Transparency b) Fairness c) Privacy d) Robustness e) Explainability

Q 11. While preparing a data set for an AI system dealing with patient data, Nora needs to determine which of the following information
i
qualifies as sensitive personal information (SPI) and needs to be removed from the data set. Which of the following would be
considered SPI?

a) Patient name b)) Marital status

c) Annual income range d) Body Mass Index (BMI)
Q 12. Imagine you’re building an AI-powered
powered tutor for students. To personalize learning experiences, the system needs access to some
tudent data. However, you prioritize protecting student privacy and don’t want to use anything that could be considered personal
pers
nformation (PI).
Which of the following would be considered personal information (PI)?

a) Home address b) Age range

c) Learning pace d) Favourite subject

Q 13. Lester is working on an AI team that is concerned with making sure that all the people in his company receive equitable treatment.
In which of the following areas of AI ethics is Lester’s team working?

a) Transparency b) Fairness c) Privacy d) Robustness e) Explainability

Q 14. Which of the following terms describes systematic errors in AI systems that, intentionally or not, generate unfair decisions
deci

a) Bias b) Partiality c) Discrimination d) Prejudice

Q 15. The first x-ray
ray in the following graphic shows no disease. An adversary adds noise to the data and then sends the data through
thr
the AI model, which results in an x-ray
ray showing disease present.
Which of the following areas of AI ethics is involved in defending against such an attack?

a) Robustness b) Fairness c) Transparency d) Privacy e) Explainab

 IBM SkillsBuild

Artificial Intelligence Fundamentals

Course - 6
Your future AI: The Job Landscape
 1. What is the primary job role of a business intelligence developer?

a) To use data tools and programming frameworks to create production-ready

production and scalable data science models
b) To analyze complex data, looking at business and market trends to increase profitability of their organization
c) To build mechanical devices that can perform tasks with commands from humans
d) To develop and maintain software that data scientists and architects use

2. Which AI professional specializes in human language and works on voice assistants and speech recognition?
a) Data scientist b) Business intelligence developer
c) Natural language processing engineer d) Software engineer

Q 3. You’re applying for an entry-level

level job as an AI professional. The interviewer asks you about your workplace skills. Which of
o the followin
skills should you elaborate on?
a) Deep learning and machine learning b) Probability and statistics
c) Programming languages, such as C++ and Java d) Communication and teamwork
4. Which of the following statements is true about industries in which AI professionals work?
a) Industries are not hiring AI professionals because AI technology is not impacting jobs
b) Only a few cutting-edge
edge industries have hired AI professionals.
c) Most industries have already hired or are in the process of hiring AI professionals
d) No industry has hired AI professionals because the industry is not yet stable.

5. You’re in a job interview and you are asked to describe your advanced AI technical skills. What set of skills should you elaborate
on?
a) Written communication, problem solving, and creative thinking
b) Big data, probability, and statistics
c) Linear algebra, problem solving, and statistics
d) Python, Tens or Flow, and machine learning