ARTIFICIAL INTELLIGENCE (SUBJECT CODE - 417)

Class X (2023-24)
Sample Paper -2 (Answer Key)

SECTION A: OBJECTIVE TYPE QUESTIONS

Q. 1 Answer any 5 out of the given 6 questions (1 x 5 = 5 marks)
i. c. Being aware of your personality 1
ii. c. Files can be of different types. 1
iii. c. Assertion is true, but Reason is false. 1
iv. d. Strengthen the global partnership for sustainable development 1
v. a. Linguistic barrier 1
vi. b. Financial stress 1

Q. 2 Answer any 4 out of the given 6 questions on Employability Skills (1 x 4 = 4 marks)

i. b. Both Assertion and Reason are true, but Reason is not the correct explanation of 1
Assertion.
ii. d. AI Ethics 1
iii.
iv.

v.
vi.
ps
c. Statement1 is correct but Statement2 is incorrect
b. Start with a medium neural network and invest in some accelerators for improved
compute capacity.
b. Chatbot
b. Rare word
1
1

1
1

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

i. b. i and iii 1
Ki
ii. a. Training data 1
iii. c. Creating digital animations for movies 1
iv. a. Edge detection 1
v. b. AI-powered conversational agent 1
vi. c. Recall and Precision 1

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

i. b. Artificial Neural Network 1
ii. c. Statement 1 is correct but Statement 2 is incorrect 1
iii. d. In an ideal scenario, half of the available data should be used for training, and the 1
other half for validation and testing.
iv. b. Convolutional Neural Network (CNN) 1

v. d. Named Entity Recognition 1

© Kips Learning Pvt. Ltd 2024

 vi. a. Overfitting 1

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

i. b. Natural Language Processing 1
ii. The evaluation metric that focuses on the proportion of actual positives that are 1
correctly identified is Recall (also known as Sensitivity or True Positive Rate).
iii. b. Resolution 1
iv. c. Dictating the proper arrangement of words in a sentence 1
v. c. Tokenization 1
vi. a. Metal detector at an airport 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 crucial for learning and improvement. It provides insights into performance, 2
offers guidance for corrections, and helps to align expectations and goals for better
outcomes. ps
Q. 7 1. Deep Breathing: Involves taking slow, deep breaths to reduce heart rate and promote 2
a sense of calm, aiding in stress reduction.
2. Mindfulness Meditation: Focuses on being present in the moment, acknowledging
thoughts and feelings without judgment, helping to alleviate stress and anxiety.
Q. 8 1. Install Antivirus Software: Anand should install reliable antivirus software and 2
regularly update it to protect his computer against malware and viruses.
2. Regular Updates: Keeping the operating system and all applications up to date
ensures that security patches are applied, reducing vulnerability to attacks.
Q. 9 An entrepreneur contributes to society by creating businesses, generating employment, 2
Ki
and fostering innovation. This drives economic growth and often leads to societal
improvements. Conversely, society provides the entrepreneur with resources, market
opportunities, and a framework of laws and norms within which to operate, influencing
the entrepreneur's success.
Q. 10 1. Environmental Stewardship: Emphasizes preserving natural resources and ecosystems 2
for future generations, promoting practices that minimize environmental impact.
2. Social Equity: Focuses on fair distribution of resources and opportunities, ensuring all
community members have access to basic necessities and rights.

Answer any 4 out of the given 6 questions in 20 – 30 words each (2 x 4 = 8 marks)

Q. 11 Social media platforms analyse user activities and preferences using AI and machine 2
learning algorithms to curate personalized feeds and targeted advertisements.

Q. 12 Priya and Navin should use an unsupervised learning model, like clustering algorithms, to 2
categorize news articles into topics without predefined categories based on inherent
data patterns.
Q. 13 1. Data Quality: Ensure accuracy, completeness, and reliability of the data. 2

© Kips Learning Pvt. Ltd 2024

 2. Data Consistency: Maintain uniform formats and standards across various sources.
Q. 14 Grayscale images represent different shades of grey, using a single intensity value per 2
pixel, whereas RGB images use three colour channels (Red, Green, Blue) per pixel.
Q. 15 ‘Even’ and ‘unless’ should not be removed as they are key conjunctions that provide 2
context and meaning to the sentence's conditional structure.
Q. 16 Based on the given information, the confusion matrix is structured as follows: 2
Predicted Positive Predicted Negative
Actual Positive 125 150
Actual Negative 45 80

Answer any 3 out of the given 5 questions in 50– 80 words each (4 x 3 = 12 marks)

Q. 17 Neural networks in artificial intelligence are like the brain's network of neurons. They 4
process information by learning patterns from data. For example, imagine teaching a
child to recognize a cat. You show them various pictures, and over time, they learn to
identify a cat regardless of its colour or size. Similarly, a neural network learns from many
examples to perform tasks like recognizing faces in photos or translating languages,
becoming better as it processes more data.
Q. 18
ps
(a) Training Dataset: This is a collection of data used to train the machine learning
model. It should be representative of the real-world scenario the model will address, and
diverse enough to allow the model to learn the necessary patterns and behaviours.
(b) Algorithm Selection Criteria: When choosing an algorithm, consider factors like the
nature and size of the data, the problem type (classification, regression, etc.),
computational resources, and desired accuracy. The algorithm should align with the
4

specific requirements and constraints of the project.

Q. 19 Based on the images provided, Dataset X appears to label different audio waveforms 4
with genres such as Rock and Jazz. If a machine learning algorithm is trained on this
Ki
dataset with labelled examples of Rock and Jazz songs, it would be capable of
distinguishing between Rock and Jazz when exposed to new songs. This is supervised
learning, where the model learns from labelled data.
Dataset Y, in contrast, seems to show various patterns without explicit labels. Learning
from Dataset Y would likely require unsupervised learning, where the algorithm tries to
find structure within the data without pre-existing labels, such as clustering similar songs
together based on their features. The key difference is that supervised learning uses
labelled data to teach the algorithm, while unsupervised learning does not use labels but
instead tries to infer the natural structure present in the data.
Q. 20 Autonomous vehicles may indeed face challenges in scenarios akin to the trolley 4
problem, where moral judgments are required. Their decision-making is bound by
algorithms and predetermined rules, lacking the nuanced ethical reasoning a human
might employ. This presents a substantial hurdle in programming AI to handle such
dilemmas where each potential action has significant moral implications.
Q. 21 (i) The total number of errors made by the diagnostic tool comprises two types: 4
False Positives: where the tool incorrectly predicted patients as 'At Risk' when they
were not (30 cases).
False Negatives: where the tool incorrectly predicted patients as 'Not at Risk' when
they were actually 'At Risk' (20 cases).

© Kips Learning Pvt. Ltd 2024

 The sum of these errors gives us the total number of errors (30 false positives + 20 false
negatives = 50 errors). These errors represent instances where the diagnostic tool's
predictions did not align with the actual reality.
The total number of errors made by the diagnostic tool is 50.
(ii) For the 'At Risk' classification:
Precision measures the accuracy of the positive predictions. It is the ratio of true
positives (correctly identified 'At Risk') to all predicted positives (true positives + false
95
positives). Here, precision is  0.76, meaning 76% of the 'At Risk' predictions
(95  30)
were correct.
Recall, or sensitivity, measures the ability to find all positive instances. It's calculated as
the ratio of true positives to actual positives (true positives + false negatives). Here,
95
recall is  0.826, meaning the tool correctly identified 82.6% of all 'At Risk'
(95  20)
cases.
The F1 Score is the harmonic mean of precision and recall, providing a balance between
𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃 𝑥𝑥 𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅
the two. It is calculated as 2 x
𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃+𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅
Here, the F1 Score is 0.792, which reflects the balance between precision and recall in
the tool's performance for 'At Risk' patients. This score is closer to 1, indicating better
ps
performance.
Therefore,
Precision is approximately 0.76.
Recall is approximately 0.826.
The F1 Score is approximately 0.792.
Ki

© Kips Learning Pvt. Ltd 2024