(1)

(2) R.K. PUBLICATIONS HOUSE

 SAMPLE QUESTION PAPER
SECTION A

Q.1. Answer any 4 out of the given 6 questions on Employability Skills (1 × 4 = 4 marks)
i. ............................. is an element of communication which encodes and sends messages
through a channel. 1
ii. Which of the following is not a way to maintain positive outlook in the long run ?
(a) Physical exercise and Fresh air (b) Healthy diet
(c) Not going out for vacations (d) Organised academic life 1
iii. Rohit is a manager in an organization. Due to some bad incident a few days ago, he is
feeling upset, annoyed and hopeless. This state of his mind is categorized as :
(a) Criminal (b) Happiness
(c) Inferiority complex (d) Stress 1
iv. In the spreadsheet software, a .................... is an arrangement of cells in a vertical (standing)
manner. 1
v. ........................... is one of the most important attitudes of an entrepreneur, which is defined
as the ability to continue to do something, even when it is difficult.
(a) Decisiveness (b) Interpersonal skills
(c) Perseverance (d) Taking initiative 1
vi. Which of the following is an importance of Greenjobs?
(a) Identify more areas to dump waste (b) Creating more IT-related jobs
(c) To protest against the use of solar fuels
(d) Protecting and restoring the ecosystem 1
Q.2 Answer any 5 out of the given 6 questions on subject specific skills (1 × 5 = 5 marks)
i. ..................... is a process with a prescribed sequence of iterative steps that data scientists
follow to approach a problem and find a solution. 1
ii. Feedback is the last stage of data science methodology that helps to refine the model and
assess it for performance and impact. (State whether this is True/False) 1
iii. What is the total number of possible colours that can be represented by combining different
intensities of red, green, and blue channels (each ranging from 0 to 255) in the RGB colour
model?
(a) 256 colours (b) 102400 colours
(c) 16 million colours (d) 1 billion colours 1
(Artificial Intelligence Sample Paper)–XII (3)
 iv. A research lab is struggling to analyze huge volumes of complex data using traditional
computers, which are taking days to deliver results. They shift to a new, cutting-edge
technology that allows them to perform parallel processing and solve complex problems
in a fraction of the time. Which advanced technology are they most likely using?
(a) Cloud Computing (b) Artificial Intelligence
(c) Quantum Computing (d) Edge Computing 1
v. Which neural network architecture has significantly transformed the field of Natural
Language Processing (NLP) and is primarily used in Large Language Models (LLMs) to
understand complex language patterns in large text datasets?
(a) Recurrent Neural Network (RNN) (b) Convolutional Neural Network (CNN)
(c) Transformer (d) Variational Autoencoder (VAE) 1
vi. Which of the following shows the correct sequence of steps to create a story through data?
(a) Create narrative → Collect data → Observe relationships → Visualize data
(b) Visualize data → Collect data → Create narrative → Observe relationships
(c) Observe relationships → Visualize data → Collect data → Create narrative
(d) Collect data → Visualize data → Observe relationships → Create narrative 1
Q.3 Answer any 5 out of the given 6 questions on subject specific skills (1 × 5 = 5 marks)
i. What is the main goal of feature engineering in machine learning?
(a) To increase the size of the dataset
(b) To reduce the number of algorithms required
(c) To select, modify, or create new features to improve model performance
(d) To visualize the dataset using graphs and charts 1
ii. • It classifies pixels belonging to a particular instance. • All the objects in the image are
differentiated even if they belong to the same class. • In this image for example the pixels
are separately masked even though they belong to the same class.
Which term can you relate the above statements to?
(a) Object Detection (b) Semantic Segmentation
(c) Instance Segmentation (d) Image Classification 1
iii. What is the main goal of image preprocessing in computer vision? 1
iv. Which two methods are commonly used for processing data in analytics?
(a) Classification and Clustering (b) Normalization and Augmentation
(c) Batch Processing and Stream Processing
(d) Preprocessing and Evaluation 1
v. In a neural network, how does learning primarily occur during training?
(a) By increasing the number of input features
(b) By deleting unnecessary neurons
(c) By adjusting weights and biases using a learning rule
(d) By reducing the number of hidden layers 1
(4) R.K. PUBLICATIONS HOUSE
 Refer to the Venn diagram that illustrates the three key elements of Data Storytelling:
Narrative, Visuals, and Data. What term should replace the question mark (?) in the center,
where all three elements intersect? 1
vi.

Q.4 Answer any 5 out of the given 6 questions on subject specific skills (1 × 5 = 5 marks)
i. A data science team develops a classification model to predict whether a patient has a
certain medical condition. After testing the model on a sample dataset, the accuracy appears
high. However, the team wants to ensure that the results are not due to random variation in
the data and genuinely reflect the model’s ability to learn from patterns.
Which evaluation method should the team use to confirm that the model’s performance
is not due to chance and is statistically reliable?
(a) Data visualization (b) Confusion matrix
(c) Feature scaling (d) Statistical significance test 1
ii. Match Column A with Column B
Column A Column B

1. Edge Detection i. Measures smoothness or roughness

2. Corner Detection ii. Analyzes colour distribution

3. Texture Analysis iii. Finds where edges meet

4. Colour-Based Feature Extraction iv. Detects sharp changes in intensity

(a) 1 → iv, 2 → iii, 3 → i, 4 → ii (b) 1 → ii, 2 → i, 3 → iv, 4 → iii

(c) 1 → i, 2 → ii, 3 → iii, 4 → iv (d) 1 → iii, 2 → iv, 3 → ii, 4 → i 1
iii. Which of the following examples best represents semi-structured data?
(a) Customer information, transaction records, product directories
(b) Audio files, images, emails, PDFs
(c) XML files, JSON files, HTML files (d) Financial reports in tabular format 1

(Artificial Intelligence Sample Paper)–XII (5)

 iv. Name the two neural networks used in a Generative Adversarial Network (GAN). 1
v. An Artificial Neural Network (ANN) with two or more hidden layers is known as a
................... . 1
vi. Which technology raises concerns about the authenticity of digital content and the spread
of misinformation?
(a) Deepfake AI (b) Generative Adversarial Networks (GANs)
(c) Variational Autoencoders (VAEs) (d) Convolutional Neural Networks (CNNs) 1
Q.5 Answer any 5 out of the given 6 questions on subject specific skills (1 × 5 = 5 marks)
i. Reva is a high school student working on a project to predict students’ final grades based
on their attendance, homework scores, and test performance. She builds a machine learning
model and evaluates it on the same data she used to train it. The model shows 98% accuracy.
Her mentor suggests that this may not reflect the model’s true performance on new data
and recommends a better technique. So, Reva splits the dataset into several parts. She
trains the model on some parts and tests it on the remaining ones. She repeats this process
multiple times to get a more reliable performance score.
Which model evaluation technique is Reva using to get a more accurate understanding
of her model’s performance?
(a) Confusion Matrix (b) Cross Validation
(c) Hyperparameter Tuning (d) Feature Engineering 1
ii. ............. is the final stage of computer vision that empowers computer vision systems to
extract valuable insights and drive intelligent decision-making in various applications. 1
iii. Identify the odd one out from the following activation functions:
(a) Sigmoid Function (b) Tanh Function
(c) ReLU (Rectified Linear Unit) (d) Linear Regression 1
iv. Name any one ethical or societal concern associated with the use of Neural Networks. 1

v.

What will come in place of ‘1’ and ‘2’? 1

vi. Assertion (A) : Characters are essential in a story as they perform actions and drive the
narrative forward.
Reason (R) : Plot refers to the events of the story, and setting refers to the time and place
where these events occur.
(a) Both A and R are true, and R is the correct explanation of A
(b) Both A and R are true, but R is not the correct explanation of A
(c) A is true, but R is false (d) A is false, but R is true 1

(6) R.K. PUBLICATIONS HOUSE

 SECTION B

Answer any 3 out of the given 5 questions on Employability Skills (2 × 3 = 6 marks)

Answer each question in 20 – 30 words.

Q.6 ‘Receiving’ is the first stage of active listening. Name the other four stages. 2
Q.7 Give any two characteristic traits of a person with Dependent Personality Disorder. 2
Q.8 Jahanvi is working on a spreadsheet. She wants to calculate the sum of cells B2 and C2, in
the cell D2. For this she has entered the formula as shown in the following image. But,
instead of the sum, she is getting the text B2 + C2 in the cell D2.

(a) What correction should be made in the cell D2 so as to get the desired sum?
(b) If the corrected formula is copied from cell D2 to D3, what value will it show in D3? 2
Q.9. Mention any four most common environmental barriers which an entrepreneur may face
while running his/her venture. 2
Q.10. Name any two Green jobs in the following sectors :
(a) Solid Waste Management (b) Agriculture 2
Answer any 04 questions out of the given 06 questions on subject
specific skills (2 × 4 = 8)

Q.11. Differentiate between : Descriptive modelling and Predictive modelling. 2

Q.12. Explain four major challenges faced by computer vision systems in real-world applications.
2
Big Data Analytics has emerged as a result of several global technology trends.
Q.13. Identify any two such trends and explain how each has contributed to the growth of Big
Data Analytics.
Q.14. What are the following known as in the given neural network?

(Artificial Intelligence Sample Paper)–XII (7)

 1. (i) (X1, X2, X3) (ii) b
2. Write the formula for calculating the predicted output.
Q.15. Differentiate between: Generative AI and Discriminative AI. 2
Q.16. Define insight in storytelling and briefly explain its significance. 2
Answer any 03 questions out of the given 05 questions on subject
specific skills (4 × 3 =12)

Q.17. Explain Confusion Matrix with a neat diagram. 4

Q.18. The following terms are related to the characteristics of Big Data:
(a) Accuracy (b) Semi-structured data
(c) Gigabytes (d) 40,000 queries per second
Identify the Big Data characteristic represented by each term and briefly explain it. 4
Q.19. Explain Recurrent Neural Networks (RNNs) with a neat diagram. List any two realworld
applications of RNNs. 4
Q.20. What are Variational Autoencoders (VAEs), and how are they different from other generative
models? Mention any two of their applications. 4
Q.21. Explain the last four stages of Freytag’s Pyramid, a classic narrative structure. 4

Answer Key
SECTION A
Ans.1. i. Encoding/Encoder
ii. (c) Not going out for vacations
iii. (d) Stress iv. columns v. (c) Perseverance
vi. (d) Protecting and restoring the ecosystem
2. i. Data Science Methodology
ii. True
iii. (c) 16 million colours
iv. (c) Quantum Computing
v. (c) Transformer
(8) R.K. PUBLICATIONS HOUSE
 vi. (d) Collect data → Visualize data → Observe relationships → Create narrative
3. i. (c) To select, modify, or create new features to improve model performance
ii. (c) Instance Segmentation
iii. To enhance the quality of the acquired image.
iv. (c) Batch Processing and Stream Processing
v. (c) By adjusting weights and biases using a learning rule
vi. change
4. i. (d) Statistical significance test
ii. (a) 1 → iv, 2 → iii, 3 → i, 4 → ii
iii. (c) XML files, JSON files, HTML files
iv. Generator and Discriminator
v. Deep Neural Network
vi. (a) Deepfake AI
5. i. (b) Cross Validation
ii. High-level processing
iii. (d) Linear Regression
iv. Data privacy, algorithmic bias, and job displacement
v. ‘1’ : User prompt and ‘2’ : Training data
vi. (b) Both A and R are true, but R is not the correct explanation of A

SECTION B

6. Understanding, Remembering, Evaluating, Responding

7. People with such a disorder are characterised by lack of self-confidence and an extra need
to be looked after. They need a lot of help in making everyday decisions and surrender
important life decisions to the care of others. They are heavily dependent on other people
for their emotional and physical needs, and thus, usually, avoid being alone.
8. (a) To perform the calculation, you must begin with =.
(b) Value in D3 will be : 49 The sum of the values in cells B3 and C3.
9. • Lack of adequate resources or raw material, • Non-availability of skilled labour. • Lack of
requisite machinery and other infrastructure, • Unavailability of monetary resources on
time.
10. (a) Green jobs related to waste management are in areas like e-waste recycling, solid waste
management, waste reduction, waste auditing, waste control, etc.
(b) Some green jobs in the agriculture sector are in organic farming, integrated pest
management, farm mechanisation and agriculture tourism.

(Artificial Intelligence Sample Paper)–XII (9)

 11. Descriptive Modeling Predictive Modeling
To summarize and understand the To use historical data and algorithms to
characteristics of a dataset without making predict future outcomes or trends.
predictions or decisions.
Describes “what is happening” in the data. Forecasts “what might happen next” based
on patterns.

Provides insights, summaries, and visual Produces predictions or estimates for

representations of existing data. unknown or future values.

Summary statistics (mean, median, mode, Regression, classification, and time-series

standard deviation, variance, range, forecasting.
percentiles) and visualizations (bar chart,
histogram, pie chart, box plot, scatter plot).

12. 1. Reasoning and Analytical Issues : Computer vision relies on more than just image
identification; it requires accurate interpretation. Robust reasoning and analytical skills
are essential for defining attributes within visual content. Without such capabilities,
extracting meaningful insights from images becomes challenging, limiting the effectiveness
of computer vision systems.
2. Difficulty in Image Acquisition : Image acquisition in computer vision is hindered by
various factors like lighting variations, perspectives, and scales. Understanding complex
scenes with multiple objects and handling occlusions adds to the complexity. Obtaining
high-quality image data amidst these challenges is crucial for accurate analysis and
interpretation.
3. Privacy and Security Concerns : Vision-powered surveillance systems raise serious
privacy concerns, potentially infringing upon individuals’ privacy rights. Technologies like
facial recognition and detection prompt ethical dilemmas regarding privacy and security.
Regulatory scrutiny and public debate surround the use of such technologies, necessitating
careful consideration of privacy implications.
4. Duplicate and False Content : Computer vision introduces challenges related to the
proliferation of duplicate and false content. Malicious actors can exploit vulnerabilities in
image and video processing algorithms to create misleading or fraudulent content. Data
breaches pose a significant threat, leading to the dissemination of duplicate images and
videos, fostering misinformation and reputational damage.
13. 1. Moore’s Law : The exponential growth of computing power as per Moore’s Law has
enabled the handling and analysis of massive datasets, driving the evolution of Big Data
Analytics.
2. Mobile Computing : With the widespread adoption of smartphones and mobile devices,
access to vast amounts of data is now at our fingertips, enabling real-time connectivity and
data collection from anywhere.
(10) R.K. PUBLICATIONS HOUSE
 3. Social Networking : Platforms such as Facebook, Foursquare, and Pinterest facilitate
extensive networks of user-generated content, interactions, and data sharing, leading to
the generation of massive datasets ripe for analysis.
4. Cloud Computing : This paradigm shift in technology infrastructure allows organizations
to access hardware and software resources remotely via the Internet on a pay-as-you-go
basis, eliminating the need for extensive on-premises hardware and software investments.
14. What are the following known as in the given neural network?
1. (i) (X1, X2, X3) Input features
(ii) b bias
2. y = w1x1+ w2x2 + w3x3 + (1 × b)

15.

16. An insight in data storytelling refers to a valuable and meaningful observation or

understanding derived from analyzing data. It involves uncovering patterns, trends,
correlations, or anomalies within the data that provide actionable information or shed light
on a particular issue or problem. Insights in data storytelling help stakeholders gain a deeper
understanding of the underlying data and empower them to make informed decisions or
take appropriate actions based on the findings.
17. A Confusion Matrix is a table used to evaluate the performance of a classification model. It
summarizes the predictions against the actual outcomes. It creates an N × N matrix, where
N is the number of classes or categories that are to be predicted. Suppose there is a problem,
which is a binary classification, then N = 2 (Yes/No). It will create a 2 × 2 matrix.
• True Positives : It is the case where the model predicted Yes and the real output was
also yes.
• True Negatives : It is the case where the model predicted No and the real output was
also No.
• False Positives : It is the case where the model predicted Yes but it was actually No.
• False Negatives : It is the case where model predicted No but it was actually Yes.
(Artificial Intelligence Sample Paper)–XII (11)
 18. (a) Accuracy
Veracity : Veracity is a characteristic in Big Data related to consistency, accuracy, quality,
and trustworthiness. Not all data that undergoes processing holds value. Therefore, it is
essential to clean data effectively before storing or processing it, especially when dealing
with massive volumes. Veracity addresses this aspect of big data, focusing on the accuracy
and reliability of the data source and its suitability for analytical models.
(b) Semi-structured data
Variety : Big data encompasses data in various formats, including structured, unstructured,
semi-structured, or highly complex structured data. These can range from simple numerical
data to complex and diverse forms such as text, images, audio, videos, and so on. Storing
and processing unstructured data through RDBMS is challenging. However, unstructured
data often provides valuable insights that structured data cannot offer. Additionally, the
variety of data sources within big data provides information on the diversity of data.
(c) Gigabytes
Volume : Every day a huge volume of data is generated as the number of people using
online platforms has increased exponentially. Such a huge volume of data is considered
Big Data. Typically, if the data volume exceeds gigabytes, it falls into the realm of big data.
This volume can range from petabytes to terabytes or even exabytes, based on surveys
conducted by various organizations. According to the latest estimates, 328.77 million
terabytes of data are created each day.
(d) 40,000 queries per second
Velocity : It refers to the speed at which data is generated, delivered and analyzed. In the
present world, where millions of people are accessing and storing information online, the
speed at which the data gets stored or generated is huge.
For example : Google alone generates more than
40,000 search queries per second. See the statistics in
the picture provided.
19. RNNs are designed for sequential data and feature
feedback loops to allow information persistence across
time steps. They have feedback connections that allow
data to flow in a loop. If the prediction is wrong, the
(12) R.K. PUBLICATIONS HOUSE
 learning rate is employed to make small changes. Hence, making it gradually increase
towards making the right prediction during the back propagation.
Application : Used in NLP for language modeling, machine translation, chatbots, as well
as in speech recognition, time series prediction, and sentiment analysis.
20. VAEs, or Variational Autoencoders, are computer programs designed to learn from data
uniquely. They consist of two parts: an encoder, which understands the data and converts it
into a hidden space called a latent space (latent space is the compressed representation of
the actual data), and a decoder, which translates the information back from this hidden
space into its original form. Unlike some other similar programs, like GANs, VAEs focus
on capturing the underlying patterns of the data to generate new samples. They find
applications in various tasks such as Data generation, detecting anomalies in data, and
filling in missing information. Both GANs and VAEs are powerful generative models but
have distinct strengths. GANs are excellent for creating visually realistic outputs, while
VAEs are better suited for structured data generation and tasks requiring interpretable latent
spaces.
21. 1. Freytag’s Pyramid is a narrative structure with five stages: exposition, rising action,
climax, falling action, and resolution. The last four stages are:
2. Risingaction : The series of events that build up to the climax of the story.
3. Climax : The most intense or important point within the story. It is often an event in
which the fortune of the protagonist turns for the better or worse in the story.
4. Fallingaction : The rest of the events that unravel after the main conflict has occurred,
but before the final outcome is decided.
5. Conclusion : The conclusion of the story where all of the conflicts are resolved and
outstanding details are explained.

(Artificial Intelligence Sample Paper)–XII (13)