SAMPLE QUESTION PAPER

ARTIFICIAL INTELLIGENCE (417)

MARKING SCHEME FOR CLASS IX (Session 2024-25)

Time Allowed: 2 Hours Maximum Marks: 50

SECTION A: OBJECTIVE TYPE QUESTIONS

Q. 1 Answer any 4 out of the given 6 questions on Employability Skills (1 x 4 = 4 marks)

i. b) Share 1
ii. c) Personal hygiene and grooming 1
iii. b) Social Networking 1
iv. a) Protocol 1
v. b) Employment generation 1
vi. a) Avoiding disposable items and plastic bags 1

Q. 2 Answer any 5 out of the given 6 questions (1 x 5 = 5 marks)

i. AI Project Cycle Mapping 1

ii. a) ROC curve 1
iii. b) Decision trees 1
iv. c) Word 1
v. d) Deep learning 1
vi. a) Resolution 1

Q. 3 Answer any 5 out of the given 6 questions (1 x 5 = 5 marks)

i. Data security 1
ii. b) It prevents data breaches that could lead to misuse of sensitive 1
information.
iii. c) Rapid technological advancements are increasing the frequency of 1
cyberattacks.
iv. Multi-Factor Authentication (MFA) 1

[1]
 v. d) Encryption 1
vi. Computer Vision 1

Q. 4 Answer any 5 out of the given 6 questions (1 x 5 = 5 marks)

i. a) Randomly scattered elements with no discernible order. 1

ii. b) A predictable change in size, colour, or position of elements within the 1

image.

iii. d) It is key for algorithms involving decision-making, natural language 1

processing, and computer vision.

iv. c) Hardware acceleration 1

v. a) Finding patterns and fixing missing or incorrect information 1

vi. b) To collect, explore, and analyse data 1

Q. 5 Answer any 5 out of the given 6 questions (1 x 5 = 5 marks)

i. d) Supervised learning is when we train the machine using unlabelled data 1

ii. b) Discriminative 1

iii. a) random noise dataset 1

iv. c) Artbreeder 1

v. d) Program plant 1

vi. a) Both A and R are true and R is the correct explanation for A 1

SECTION B: SUBJECTIVE TYPE QUESTIONS

Answer any 3 out of the given 5 questions on Employability Skills (2 x 3 = 6 marks)
Answer each question in 20 – 30 words.
Q. 6 Feedback is important in communication because it: 2
• Clarifies understanding: Feedback helps ensure that the message has
been correctly understood by the recipient, preventing misunderstandings
and fostering clear communication.
• Builds trust: By providing feedback, communication can become more
effective, creating a sense of trust and rapport between the sender and
recipient.

[2]
Q. 7 In SMART goals, "R" stands for "Relevant". A goal is relevant if it aligns with 2
the larger objectives of the organisation or individual. For example, if a
company’s goal is to expand market reach, a relevant goal might be to develop
targeted marketing strategies that directly support that objective.
Q. 8 Two methods to protect our data on the computer: 2
• Encryption: Converts your data into a coded form that can only be read
with the correct decryption key. This protects your data when it's stored
on your device or transmitted over the internet. You can encrypt files or
use full disk encryption to prevent unauthorised access.
• Back up your data: Regularly back up your data to a secure external
storage device that's kept somewhere other than your main workspace. If
possible, encrypt the storage device as well.
Q. 9 Entrepreneurs perform many roles when running a business, including: 2
• Identifying opportunities: Entrepreneurs identify unmet market needs
and develop solutions to create value for their target audience.
• Managing and organising: Entrepreneurs create business plans, hire
employees, and allocate their responsibilities. They also organise and
oversee the business structure and environment.
Q. 10 Raj is in the "Survive" stage of the entrepreneurial career process. In this 2
stage, he faces competition but manages to sustain his business despite
competing convenience stores in the area, demonstrating resilience in a
competitive market.
Answer any 4 out of the given 6 questions in 20 – 30 words each (2 x 4 = 8 marks)
Q. 11 The main ethical principles guiding AI development are: 2
• Human Rights: Ensuring AI respects fundamental human rights.
• Bias: Striving to prevent and minimise biases within AI systems.
• Privacy: Protecting individual privacy when handling personal data.
• Inclusion: Promoting inclusivity, ensuring AI benefits everyone.
Q. 12 Data literacy can equip individuals with skills and knowledge to improvise in a 2
tech and data-driven world. The following are the reasons why data literacy is
critical to an organisation:
• Data literacy enhances decision-making ability in individuals based on
evidence.
• Data literacy is able to cultivate critical thinking skills to understand and
explore data’s implications
• Data literacy helps in analytically producing solutions to problems
• Data literacy fuels innovation by providing tools and techniques to explore
data from different perspectives.

[3]
Q. 13 Probability is a branch of statistics that deals with the likelihood or chance of 2
an event to happen or different outcomes occurring in a given situation.
• It measures the amount of certainty of an event.
• It helps us make predictions about future events based on the data we
have.
• It helps us understand how likely something is to happen.
The formula for probability is given as the ratio of the Number of Favourable
Outcomes to the Total Number of Possible Outcomes.
P(A) = Number of Favourable Outcomes / Total Number of Possible Outcomes
Q. 14 The most possible reasons why data security is more important now are: 2
• A constant fear cyberattacks affect all people.
• The fast-technological changes will boom cyberattacks.
• A persistent fear everyone is impacted by cyberattacks.
• Rapid technical advancements will increase the frequency of cyberattacks.
Q. 15 A system map is a diagrammatic representation of a set of things working 2
together. It focuses on the components and boundaries of a system. System
map helps us to find relationships between different elements of the problem
which we have scoped. It helps to find a solution to achieve the goal of our
project.
Q. 16 Unsupervised learning is a type of machine learning where models are trained 2
using data that does not have labels. This means the model has to find
patterns and relationships in the data on its own.
Generative modelling is a specific approach within unsupervised learning that
focuses on understanding and modelling how the data is generated.
Answer any 3 out of the given 5 questions in 50 – 80 words each (4 x 3 = 12 marks)

Q. 17 Different types of data security controls are as follows: 4

• Strong Passwords: A strong password is a combination of atleast 8
characters with upper and lower-case letters, numbers, and special
characters that is difficult for unauthorised individuals or automated
programs to guess or crack. It is a very basic step that one should take and
never share the same with even the most trusted ones. Avoid using birth
dates, anniversary dates, and common combinations of numbers.
• Authentication: It also refers to multi-factor authentication (MFA) and is
an additional security layer in online data systems. After a user enters their
password to log in, MFA requires them to provide one or more additional
forms of authentication to verify their identity. This could include onetime
generated code as a security token in smartphones or emails or a
fingerprint or facial recognition, or Passwords or PINs.

[4]
 • Access Controls: Access controls refer to the security measures and
protocols to restrict access to sensitive data, ensuring that only authorised
individuals or entities can view, modify, or interact with it. This reduces
the risk of unauthorised access by limiting the number of users who can
interacts with sensitive data.
• Data Backup: Data backup refers to the process of creating copies of data
to ensure that it can be restored in the event of data loss due to natural
disasters, accidents, cyber-attacks, or other unexpected events.
Sometimes physical backup media is used to secure in access-controlled
environments. Another method to secure data can be the cloud backup
which is considered more reliable.
Q. 18 The data literacy framework provides a comprehensive and structured 4
approach to develop the necessary skills for using data efficiently and with all
levels of awareness. Each level builds upon the previous one, fostering a
deeper and more understanding ability to work with data.
Level 1: Plan: Planning sets a clear roadmap and structured approach to
enhance data literacy across different levels of awareness. You can:
• define the specific and measurable goals.
• develop a timeline and milestones for achieving these goals.
• identify and allocate resources needed (e.g., budget, tools, personnel).
Level 2: Communication: Clear and consistent communication about data
literacy will ensure an efficient data literacy framework within an
organisation.
Level 3: Assess: Introducing participants to data literacy assessment tools and
finding out how comfortable they are with data is crucial in data literacy
program.
Level 4: Develop Culture: To integrate data literacy skills into the
organisational culture, it is important to make data-driven decision-making a
fundamental part of everyday work. This
Level 5: Prescriptive Learning: It refers to an educational approach that
focuses on guiding students towards specific outcomes or behaviours through
structured instruction and targeted interventions based on data and analysis
of their learning needs.
Level 6: Evaluate: Designing an evaluation metric for the data literacy
program involves creating a structured framework to assess participants’
progress and the effectiveness of the program overall.
Q. 19 Generative Artificial Intelligence (AI), also known as Gen AI, encompasses 4
algorithms designed to produce new data closely resembling human-
generated content across various forms, including audio, code, images, text,
simulations, and videos. Leveraging existing data and content for training,

[5]
 Generative AI holds promise for applications spanning natural language
processing, computer vision, the metaverse, and speech synthesis. As it
undergoes continuous training with more data, this technology evolves,
progressively refining its capabilities. Its growing popularity stems from the
intuitive interaction it enables, allowing users to prompt the AI using natural
language.
Generative AI comes in a variety of forms, each with unique advantages and
uses. Some of the most typical varieties are as follows:
• Generative Adversarial Networks (GANs): Two neural networks
(generator and discriminator) work in opposition to create realistic data.
• Variational Autoencoders (VAEs): Generate data by encoding and
decoding, capturing latent representations for diverse outputs.
• Recurrent Neural Networks (RNNs): Often used in sequential data
generation, such as text and music, due to their ability to handle time-
series information.
• Autoencoders (AEs): Compress and reconstruct data, useful in generating
content and reducing noise in images.
Q. 20 • Certain Events: These events are guaranteed to happen; there is no doubt 4
about their occurrence. It will have a probability of 1. For example:
o If you flip a fair coin, the probability of it landing heads up or tails up is
certain, as one of these outcomes is guaranteed.
o The occurrence of sunrise and sunset each day is certain.
o When you click a light switch, the light bulb will either turn on or off.
o The act of inhaling and exhaling is certain as long as a person is alive.
o Time consistently moves forward, and each passing moment is certain.
o The beating of the heart is a certain event, as long as a person is alive.
o If it is Sunday today, it is certain, tomorrow is going to be a Monday.
• Likely Events: These events have a higher probability of occurring as
compared to other events. For example:
o If you roll a fair six-sided die, the likelihood of rolling a number greater
than 2 (3, 4, 5, or 6) is higher than rolling a number less than or equal
to 2 (1, or 2).
o If you visit a store known for carrying a wide variety of brands, it’s likely
that you’ll find your favourite brand among their products.
o If you study well for an exam, you’re more likely to pass with good
grades.
o If you buy many raffle tickets compared to others, you’re more likely
to win a prize.

[6]
Q. 21 4
Factor AE VAE
Basic Function Neural network model that Similar to AE but
learns to encode input data incorporates probabilistic
into a compressed elements to learn a latent
representation and then space representation of
decode it back to the input data.
original data.
Latent Space Deterministic, fixed- Probabilistic, continuous
Representation dimensional encoding of latent space
input data. representation, allowing
for sampling of data points.
Factor AE VAE
Reconstruction Minimises the difference Same as AE but also
Loss between the input data includes a regulariser to
and its reconstructed enforce a Gaussian-like
output. distribution in the latent
space.
Handling Can suffer from overfitting Less prone to overfitting
Overfitting due to the fixed encoding due to the probabilistic
structure. nature of the latent space,
which allows for smoother
generalisation.

[7]