Section A - Objective Type Questions
i.Which of the following is an effective component of good feedback?
a) Detailed and time-consuming
c) Specific
b) Indirect
d) Opinion-based
✅c) Specific
ii.Which of the following is an example of self-motivation?
(a) Waiting for someone to tell you what to do
(b) Complaining about problems
(c) Taking initiative to learn a new skill
(d) Ignoring responsibilities
✅(c) Taking initiative to learn a new skill
iii.CBSE and AIIMS recently launched a mental health awareness series for students and teachers.
Which of the following practices aligns most closely with effective self-management in terms of mental well-being?
a) Ignoring emotional stress to maintain academic focus
b) Practicing regular reflection, time-blocking for rest, and seeking help when overwhelmed
c) Comparing oneself constantly to high-performing peers
d) Multitasking throughout the day without prioritizing mental breaks
✅b) Practicing regular reflection, time-blocking for rest, and seeking help when overwhelmed
iv.Tanvi stores her school project on Google Drive so she can access it from both school and home. She also shares the file with her teacher via a shared link. Which ICT skill is being used here?
a) Using spreadsheet formulas
b) Managing offline backups
c) Using cloud-based storage and collaboration tools
d) Installing new operating systems
✅c) Using cloud-based storage and collaboration tools
V.Entrepreneurs help in reducing.
(a) GDP
(b) Unemployment
(c) Jobs
(d) Innovation
✅(b) Unemployment
vi.A factory is using a large amount of fresh water to cool its machinery and then discharging hot, used water into a nearby river. A new policy is introduced requiring them to treat the water and use it in a closed-loop system to reduce overall water consumption. This policy encourages a green skill related to:
a) Recycling water for irrigation
b) Practicing water harvesting
c) Adopting a circular water management approach
d) Using chemical treatments to reduce water temperature
✅c) Adopting a circular water management approach
Q2.Answer any 5 out of the given 6 questions.
i.academics, especially in science, which is his favourite subject. Assertion (A): Raman is not good at academics; however, he is a great athlete and often wins prizes in inter-school or even inter-state matches. Gaurav is not good at sports, but he excels in
Reason(R): Humans possess different types of intelligence but at different levels.
a) Both A and R are correct, and R is the correct explanation of A.
b) Both A and R are correct, but R is not a correct explanation of A.
c) A is correct, but R is not correct.
d) A is not correct, but R is correct.
✅a) Both A and R are correct, and R is the correct explanation of A.
ii.Whenever we search "nurse" on Google, it shows images of mostly women nurses. When you search for "shirt", it mostly shows shirts for men. These are examples of:
a) Al Ethics
b) Data Privacy
c) Al Bias
d) Al Access
✅c) Al Bias
iii.Statement 1: There are four layers in a neural network.
Statement 2: The first layer of the neural network is known as the output layer.
a) Both Statement1 and Statement2 are correct.
b) Both Statement1 and Statement2 are incorrect.
c) Statement1 is correct but Statement2 is incorrect.
d) Statement2 is correct but Statement1 is incorrect.
✅c) Statement1 is correct but Statement2 is incorrect.
iv.In Computer Vision O represents and 1 represents colour.
a) red and blue
c) white and black
b) black and white
d) blue and black
✅b) black and white
v.Which of the following refers to the meaning of the word in a sentence that helps in interpreting the proper message of the complete structure of words?
a) Syntax
c) Analogy
b) Semantics
d) Language
✅b) Semantics
vi.A corpus contains 4 documents in which the word 'diet' appears once in document 1. Identify the term in which we can categorize the word 'diet.
a) Stop word
b) Rare word
c) Frequent word
d) Removable word
✅b) Rare word
Q3.Answer any 5 out of the given 6 questions.
i.An Al system uses two broad classes of data namely content data which includes the raw video streams title, description, etc, and user activity data that includes rating a video, favoriting/liking a video, or subscribing to an uploader, and watch time. Based on this, the Al system measures a user's engagement and happiness. It then starts computing personalized recommendations for the user. Which of the following applications can you relate to this?
(a) Self-driving car
(b) email filters
(c) Siri
(d) YouTube
✅(d) YouTube
ii.The......... canvas helps you in identifying the key elements related to the problem.
(a) Problem scoping
b) Project cycle
(c) 4Ws Problem
(d) Algorithm
✅(c) 4Ws Problem
iii.Statement 1: Data Science primarily works around analysing the data.
Statement 2: It employs techniques and theories drawn from many fields within the context of Mathematics, Statistics, Computer Science, and Information Science.
(a) Both Statement 1 and Statement 2 are correct.
(b) Both Statement I and Statement 2 are incorrect.
(c) Statement 1 is correct but Statement 2 is incorrect.
(d) Statement 2 is correct but Statement 1 is incorrect.
✅(a) Both Statement 1 and Statement 2 are correct.
vi.How many channels does a colour image have?
Answer.3(RGB)
RED
GREEN
BLUE
v.the sub-field of Al that is focused on enabling computers to understand and process human languages.
(a) Deep Learning
(b) NLP
(c) Machine Learning
(d) Data Science
✅(b) NLP
vi.Which of the following is defined as the measure of balance between precision and recall?
(a) Accuracy
(b) Reliability
(c) F1 Score
d) Punctuality
✅(c) F1 Score
Q4.Answer any 5 out of the given 6 questions.
i.
Assertion (A): The Al project cycle starts with the problem scoping stage.
Reason(R): Without clearly defining the problem, data collection and model building may go in the wrong direction.
(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, R is false
(d) A is false, R is true
✅(a) Both A and R are true, and R is the correct explanation of A
ii.What is the primary input for computer vision systems?
a) Text documents
b) Visual data such as images and videos
c) Numerical data
d) Audio signals
✅b) Visual data such as images and videos
iii.Which of the following is a challenge in NLP?
a) Ambiguity in human language
b) Limited computational power
c) Lack of data
d) Complexity of algorithms
✅a) Ambiguity in human language
iv.Assertion: An Al model should strike a balance between underfitting and overfitting to be a good fit.
Reasoning: Overfitting and Underfitting are the two primary causes of poor performance in Al models.
Choose the correct option:
a. Both A and R are true, and R is the correct explanation for A
b. Both A and R are true, and R is not the correct explanation for A
c. A is True, but R is False
Which of the following is NOT a component of a confusion matrix?
d. A is false, but R is True
✅
v.Which of the following is NOT a component of a confusion matrix?
d. A is false, but R is True
a) Mean Squared Error
b) False Negatives
c) True Positives
d) True Negatives
✅a) Mean Squared Error
vi.Grouping customers based on their shopping patterns into segments like Budget Shoppers or Premium Buyers is an example of:
a) Classification
b) Regression
c) Clustering
d)Association
✅c) Clustering
Q5.Answer any 5 out of the given 6 questions.
i.You would feed the data into the machine. This is the data with which the machine can be trained. Now, once it is ready, it will predict his next data efficiently. This previous data is known as.
a) Testing Data
b) Training Data
c) Exploring Data
d) All of the above
✅b) Training Data
ii.In computer vision, what does 'image segmentation' refer to?
a) Compressing an image to reduce its size
b) Enhancing the brightness of an image
c) Converting an image to grayscale
d) Dividing an image into multiple parts or regions
✅d) Dividing an image into multiple parts or regions
iii.Which feature of NLP helps in understanding the emotions of the people mentioned with the feedback?
(a) Virtual Assistants
(b) Sentiment Analysis
(c) Text classification
(d) Automatic Summarization
✅(b) Sentiment Analysis
vi.Which metric is best when the cost of false negatives is high?
a) Accuracy
b) F1 Score
c) Recall
d) Precision
✅c) Recall
v.What is the purpose of splitting data into training and testing sets?
a) To increase training speed
b) To improve model accuracy
c) To reduce the size of the dataset
d) To evaluate model performance on unseen data
✅d) To evaluate model performance on unseen data
vi.Which one of the following is the second stage of Al project cycle?
a) Data Exploration
b) Data Acquisition
c) Modeling
d) Problem Scoping
✅b) Data Acquisition
Section B- Subjective Type Questions
SHORT METHOD 👇
(a). What type of communication was Priya attempting?
Short Answer:-Verbal communication (public speaking).
(b). What could Priya do next time to feel more confident?
Short Answer:-Practice in front of a mirror or with friends, take deep breaths, and stay calm. She could also write key points on a cue card.
LONG METHOD 👇
Answer
(a) What type of communication was Priya attempting?
👉Priya was attempting verbal communication because she was speaking to an audience (morning assembly) to convey her message.
More specifically, this is public speaking, which is a form of oral communication in front of a group.
(b) What could Priya do next time to feel more confident?
👉Some practical strategies:
Practice multiple times – Rehearse the speech in front of a mirror, friends, or family.
Use cue cards – Write key points to help remember lines.
Deep breathing / relaxation – Helps reduce nervousness before speaking.
Positive visualization – Imagine giving a successful speech confidently.
Start small – Practice speaking in smaller groups before a big audience.
Engage with the audience – Make eye contact, smile, and focus on communication, not perfection.
Q7.Explain any two time management techniques?
SHORT METHOD:-
To-do list: A daily checklist of tasks to be completed.
Prioritization: Organizing tasks based on urgency and importance to use time effectively.
LONG METHOD.
1. Pomodoro Technique ⏲️
How it works:
Break work into short focused intervals (usually 25 minutes), called Pomodoros, followed by a 5-minute break.
After completing 4 Pomodoros, take a longer break (15–30 minutes).
Benefits:
Improves focus and concentration.
Prevents burnout by balancing work and rest.
Example:
Study for 25 minutes → 5-minute break → repeat 4 times → take 20-minute break.
2. Eisenhower Matrix / Urgent-Important Matrix 🗂️
How it works:
Tasks are divided into four categories based on urgency and importance:
Important & Urgent → Do immediately
Important & Not Urgent → Schedule for later
Not Important & Urgent → Delegate if possible
Not Important & Not Urgent → Eliminate or minimize
Benefits:
Helps prioritize tasks effectively.
Reduces stress and wasted time by focusing on what really matters.
Example:
Studying for exams (Important & Urgent) → do now
Planning next week’s schedule (Important & Not Urgent) → schedule
Q8.What is netiquette? Why is it important?
SHORT METHOD
Answer.Netiquette: Netiquette refers to proper and respectful behavior on the internet.
Importance of Netiquette: It ensures safe, ethical, and effective communication. online.
LONG METHOD
Answer.Netiquette is a combination of “network” + “etiquette.”
It refers to the set of rules and guidelines for polite, respectful, and responsible behavior while communicating online—through emails, chats, social media, forums, or any internet platform.
Examples of netiquette:
Using polite language in emails and chats.
Avoiding typing in ALL CAPS (it’s considered shouting).
Giving credit when sharing someone else’s work.
Avoiding spam or offensive messages.
Q9.How do entrepreneurs help in the development of a country?
SHORT METHOD
Answer.Entrepreneurs create jobs, promote innovation, increase productivity, and contribute to GDP, thereby helping in the economic development of a country.
LONG METHOD
Answer.Ways Entrepreneurs Help in National Development:
Creating Employment Opportunities 👷♂️
By starting businesses, entrepreneurs generate jobs for people, reducing unemployment.
Example: Small startups, factories, or IT firms employ thousands of workers.
Boosting Economic Growth 💰
Entrepreneurship leads to production of goods and services, increasing the country’s GDP.
Encourages investments and stimulates economic activity.
Promoting Innovation and Technology 💡
Entrepreneurs introduce new ideas, products, and services.
Example: Tech startups develop apps, software, and gadgets that improve efficiency and lifestyle.
Increasing Competition and Quality ⚖️
Competition among businesses improves product quality and reduces prices for consumers.
Generating Revenue for Government 🏛️
Businesses pay taxes which can be used for infrastructure, education, healthcare, and public services.
Encouraging Exports and Foreign Investment 🌍
Entrepreneurs help in producing goods and services that can be exported, bringing foreign exchange into the country.
Social Development 🌱
Many entrepreneurs engage in corporate social responsibility (CSR) initiatives like building schools, hospitals, and promoting skill development.
Q10.Write four development goals given by United Nations Sustainable Development Summit 2015.
SHORT METHOD
Answer.Four development goals given by the United Nations Sustainable Development Summit 2015 are: - Goal 4: Quality Education - Goal 6: Clean Water and Sanitation -Goal 7: Affordable and Clean Energy-Gool 13: Climate Action
LONG METHOD
Answer.The United Nations Sustainable Development Summit 2015 adopted the 2030 Agenda for Sustainable Development, which includes 17 Sustainable Development Goals (SDGs). Here are four important goals from that list:
Four UN Sustainable Development Goals (2015):
No Poverty 🌍
End poverty in all its forms everywhere.
Zero Hunger 🌾
End hunger, achieve food security, and promote sustainable agriculture.
Good Health and Well-being 🏥
Ensure healthy lives and promote well-being for all at all ages.
Quality Education 📚
Ensure inclusive and equitable quality education and promote lifelong learning opportunities for all.
Q11.What do you mean by Bioethics?
SHORT METHOD
Answer.Bioethics is an ethical framework in healthcare and life sciences that addresses moral issues in medicine and biology. It ensures Al and scientific advances follow ethical standards, promoting responsible and humane practices
LONG METHOD
Answer.Bioethics is the branch of ethics that deals with the moral, social, and legal issues arising from advances in biology, medicine, and life sciences.
It focuses on determining what is right or wrong in the use of biological and medical knowledge.
Key Areas of Bioethics:
Medical Ethics: Proper conduct in healthcare (e.g., patient consent, confidentiality).
Genetic Engineering: Ethical use of biotechnology, gene editing, cloning.
Research Ethics: Moral rules for experiments on humans or animals.
Environmental Ethics: Responsible use of biological resources and conservation.
Example:
Is it ethical to clone animals or humans?
Should doctors perform euthanasia (mercy killing)?
Is it right to conduct experiments on humans without consent?
Q12.Give the difference between rule-based and learning-based Al models.
SHORT METHOD
Answer.Rule-based Al Models: Work on predefined rules and logic created by humans. They cannot learn from data. Example An expert system that diagnoses a disease based on IF THEN rules. Learning-based Al Models: Learn patterns and knowledge automatically from data. Improve performance with more training. Example: A spom filter that learns from labelled emails (spam or not spam).
LONG METHOD
Answer.Difference Between Rule-Based and Learning-Based AI Models
Feature
Rule-Based AI
Learning-Based AI
Definition
AI system that follows a set of predefined rules to make decisions.
AI system that learns patterns from data to make predictions or decisions.
How it Works
Works on “if-then” rules written by humans.
Works on training data using machine learning algorithms.
Flexibility
Less flexible; cannot handle situations outside the predefined rules.
Highly flexible; can adapt to new data and situations.
Example
Expert systems, decision trees with fixed rules.
ChatGPT, self-driving car models, image recognition systems.
Pros
Easy to understand and implement; predictable behavior.
Can handle complex problems; improves with more data.
Cons
Cannot learn on its own; fails if rules are incomplete.
Q13.What is supervised, and unsupervised learning? Explain with examples.
SHORT METHOD
Answer.Supervised Learning:- Uses labelled data (input-output pairs). Example: Predicting house prices based on size, location, and labelled prices.
Unsupervised Learning:- Uses unlabelled data to find patterns/groups. Example: Grouping customers into clusters based on purchasing habits.
LONG METHOD
Answer.. Supervised Learning ✅
Definition:
Supervised learning is a type of machine learning where the model is trained on labeled data.
The model knows the correct answers for each input and learns to predict the output.
Key Points:
Requires input-output pairs.
Goal: Predict the output for new, unseen inputs.
Example:
Email Spam Detection: Emails labeled as Spam or Not Spam. The model learns from these labels to classify new emails.
Predicting House Prices: Input = house features (size, location), Output = price. The model learns from historical data to predict new prices.
2. Unsupervised Learning ✅
Definition:
Unsupervised learning is a type of machine learning where the model is trained on unlabeled data.
The model tries to find patterns, structures, or groups in the data on its own.
Key Points:
No labeled outputs are provided.
Goal: Discover hidden patterns or clusters in data.
Example:
Customer Segmentation: Grouping customers into segments like Budget Shoppers and Premium Buyers based on purchase behavior.
Market Basket Analysis: Finding which products are often bought together in a store.
Q14.Explain train-test split with an example.
SHORT METHOD
Answer.Train-test split is a method to assess a model's performance on unseen data by dividing the dataset into:
Training set-used to train the model
Testing set-used to evaluate its accuracy
Example: From 1,000 student records, use 800 (80%) for training and 200 (20%) for testing
LONG METHOD
Answer.
Train-Test Split is a technique in machine learning where the dataset is divided into two parts:
Training Set: Used to train the model, i.e., teach it patterns from data.
Testing Set: Used to evaluate the model on new, unseen data to see how well it generalizes.
This ensures the model does not just memorize the data (overfitting) and can predict accurately on new inputs.
How it Works:
Suppose we have 1000 data points:
80% (800 points) → Training set
20% (200 points) → Testing set
Step 1: Model learns from the training data.
Step 2: Model predicts on the testing data to check accuracy and performance.
Example:
Task: Predict house prices.
Dataset: 1000 houses with features (size, location, rooms) and price.
Training: Use 800 houses to train the model.
Testing: Use 200 houses to see how well the model predicts prices for houses it hasn’t seen.
Outcome: If the predictions are close to actual prices, the model is good. If not, the model may need more data or tuning.
Q15.Explain with an example how computer vision can be used in traffic management. Mention its benefits and limitations.
SHORT METHOD
Answer .In traffic management, computer vision systems use cameras to detect vehicle counts, identify violations, and adjust signal timings. Example: Automatic number plate recognition for issuing e-challans. Benefits o Real-time monitoring, o Reduced manual intervention. Limitations: o Accuracy drops in poor lighting or bad weather. o Requires high-quality cameras and maintenance.
LONG METHOD
Answer.
Computer Vision in Traffic Management
Computer vision (CV) uses cameras and AI algorithms to “see” and interpret the environment. In traffic management, CV systems can monitor roads, vehicles, and pedestrians in real-time to improve traffic flow and safety.
Example: Smart Traffic Signal Control
Imagine a busy city intersection:
Setup: Cameras are installed at intersections to capture live traffic footage.
Computer Vision Analysis:
Detects the number of vehicles waiting at each lane.
Identifies traffic congestion or accidents.
Tracks vehicle types (cars, buses, bikes) for better traffic predictions.
Action:
Traffic lights adjust automatically: green lights are extended for lanes with more vehicles.
Alerts are sent to traffic police if accidents or unusual congestion are detected.
Result: Traffic flow is smoother, reducing waiting time and congestion.
Real-world example: Some cities like Singapore and Pune (India) have implemented AI-based traffic signal systems that use CV to dynamically control traffic lights.
Benefits
Reduced Congestion: Signals adapt to real-time traffic conditions.
Improved Safety: Detects accidents, speeding, or pedestrian violations quickly.
Data Insights: Provides analytics for city planning (e.g., busiest routes, peak hours).
Cost-effective: Reduces the need for manual traffic monitoring.
Limitations
High Initial Cost: Cameras, servers, and AI software can be expensive.
Weather Dependency: Heavy rain, fog, or snow may reduce camera accuracy.
Privacy Concerns: Continuous video monitoring may raise privacy issues.
Maintenance: Cameras and sensors require regular upkeep to function properly.
Limited Understanding: AI might misinterpret unusual situations (e.g., construction detours, unusual vehicle shapes).
Q16.Mention some applications of Natural Language Processing.?
SHORT METHOD
Answer.Natural Language Processing Applications- Sentiment Analysis. Chatbots & Virtual Assistants. Text Classification. Text Extraction. Machine Translation.
Text Summarization. Market Intelligence .Auto-Correct.
LONG METHOD
Answer.Applications of NLP
Chatbots and Virtual Assistants
Examples: Siri, Alexa, Google Assistant, ChatGPT
Use NLP to understand user queries and respond in natural language.
Machine Translation
Examples: Google Translate, Microsoft Translator
Converts text from one language to another automatically.
Sentiment Analysis
Analyzes opinions in text (like reviews or social media posts) to detect positive, negative, or neutral sentiments.
Example: Companies analyzing customer feedback.
Spam Detection
Email services like Gmail use NLP to filter spam or phishing emails.
Text Summarization
Automatically generates a short summary of long articles, reports, or news.
Example: Summarizing research papers or news feeds.
Speech Recognition
Converts spoken language into text.
Example: Voice typing, transcription software.
Question Answering Systems
Finds answers from large datasets or documents.
Example: IBM Watson, ChatGPT answering questions.
Information Extraction
Extracts structured data from unstructured text.
Example: Pulling names, dates, or locations from news articles.
Text-to-Speech (TTS) & Speech-to-Text (STT)
Converts text into natural-sounding speech and vice versa.
Example: Audiobooks, accessibility tools for visually impaired.
Autocorrect and Grammar Checking
Tools like Grammarly or Microsoft Word use NLP to correct spelling and grammar mistakes.
Q17.Air pollution kills an estimated seven million people worldwide every year. WHO data shows that 9 out of 10 people breathe toxic air. From smog hanging over cities to smoke inside the home. air pollution poses a major threat to human health and the climate. The major pollution sources include vehicles, power generation, agriculture/waste incineration, and industry. Harmful gaminy like SO2, NO2, CO are emitted directly into the air because of pollution. Deploying an air quality Index monitor is one way that would help to know the local air quality and take action to protect their health. Taking this as the problem, frame the problem statement.
SHORT METHOD
answer.Problem Statement:
Air pollution is a severe global issue affecting human health and the environment. Due to harmful emissions from vehicles, industries, and power generation, people are exposed to toxic air containing gases like SO,, NO,, and CO. There is a need to design and implement an Air Quality Index (AQI) Monitoring System that can measure local air quality levels and provide real-time data to help individuals and authorities take timely actions to reduce health risks and pollution.
LONG METHOD
Answer .Problem Statement: Air Pollution Monitoring and Awareness
Air pollution is a severe global health and environmental challenge, responsible for an estimated seven million deaths each year. According to the World Health Organization (WHO), 9 out of 10 people worldwide breathe toxic air, exposing them to harmful pollutants such as SO₂, NO₂, and CO. Major sources of pollution include vehicles, power generation, agriculture/waste incineration, and industrial activities, contributing to smog in cities and indoor air contamination.
Despite its grave impact on human health and the climate, many people are unaware of the real-time quality of the air they breathe, which limits their ability to take preventive measures.
The challenge is to develop a reliable and accessible system, such as an Air Quality Index (AQI) monitor, that can measure local air pollution levels in real-time and provide actionable information to the public. This system should enable individuals and communities to take timely actions to protect their health and reduce exposure to harmful pollutants.
SHORT METHOD
Answer.The image shows two simple graphs labelled (a) and (b):
(a) displays two clusters of data points enclosed by ovals representing clustering, where similar data points are grouped.
(b) shows data points following a curved line representing regression where a model fits a line (or curve) to predict continuous values.
So, the figure illustrates:
( a) Clustering (unsupervised learning)
(b) Regression (supervised learning)
LONG METHOD
Answer.
(a) Algorithm: Clustering (K-Means Clustering)
Identification
In picture (a), data points are grouped into two separate oval-shaped groups without any labels. This shows unsupervised learning, where the algorithm groups similar data points together.
Algorithm Name
👉 K-Means Clustering
Explanation
K-Means is an unsupervised machine learning algorithm.
It divides the data into K groups (clusters) based on similarity.
Data points inside one cluster are more similar to each other than to points in other clusters.
Working Steps
Choose number of clusters K.
Randomly select K centroids.
Assign each data point to the nearest centroid.
Recalculate centroids.
Repeat until clusters do not change.
Use Cases
Customer segmentation
Image segmentation
Pattern recognition
(b) Algorithm: Regression (Non-Linear Regression)
Identification
In picture (b), points are scattered but a smooth curved line fits the data points. This shows a relationship between input and output values.
Algorithm Name
👉 Regression (Non-linear Regression)
Explanation
Regression is a supervised learning algorithm.
It is used to predict continuous values.
Since the line is curved, it represents non-linear regression.
Working
The algorithm finds the best-fit curve for the given data.
It minimizes the error between actual points and predicted values.
Use Cases
Price prediction
Growth trends
Weather forecasting.
Q19.The automated trade industry has developed an Al model that predicts the selling and purchasing of automobiles. During testing, the Al model came up with the following predictions.
SHORT METHOD
Answer:-
Reality:Yes
Reality:no
Predicted:YES TP = 60 ,FP-25,
Predicted: NO. FN=5 ,TN = 10
(a) Total Tests Performed:
Total = TP+FP + FN + TN
60+25+5+10= 100 tests
(b)Calculate Precision, Recall and F1 Score.
Precision TP/(TP+FP)
=60/(60+25)
=60/85
=0.7059 (or 70.59%)
Recall TP/(TP+FN)
=60/(60+5)
=60/65
=0.9231 (or 92.51%)
F1 Score =2x (Precision ×Recall) / (Precision + Recall)
=2x (0.7059× 0.9231)/(0.7059+ 0.9231)
=2×(0.6519)/1.629
=0.7998 (or 79.98%)
LONG METHOD
Answer.Given Confusion Matrix
Answer.Given Confusion Matrix
Predicted \ Reality
Step 1: Identify values
True Positive (TP) = 60
(Predicted Yes & Actually Yes)
False Positive (FP) = 25
(Predicted Yes but Actually No)
False Negative (FN) = 5
(Predicted No but Actually Yes)
True Negative (TN) = 10
(Predicted No & Actually No)
(a) Total number of tests performed
Total Tests =TP+FP+FN+:
=60+25+5+10=100
Total tests performed = 100
(b) Precision, Recall and F1 Score
1. Precision
Precision =TP/ TP+FP
=0.7059
Recall = TP/(TP+FN)
= 0.923
3. F1 Score
F1 Score = 2x (Precision×Recall)/Precision+Recall
=0.7998or(approx
Final Answer Summary (For Exam)
Total Tests = 100
Precision = 0.71 (71%)
Recall = 0.92 (92%)
F1 Score = 0.80 (80%)
Q20.Case Study:
Al in Healthcare-X-ray Analysis A hospital uses a computer vision-powered Al tool to analyze chest X-rays for early detection of pneumonia. The system highlights suspicious areas for doctors to review, helping them make faster diagnoses.
(a)Which type of neural network is most suitable for analyzing X-ray images?
a) RNN
b) CNN
c) GAN
d) SVM
(b) Highlighting suspicious areas in an X-ray is an example of:
a) Image segmentation
b) Image compression
c) Image encryption
d) Image synthesis
(c) Why is it important for doctors to review Al-generated results instead of relying solely on
them?
(d) Discuss one ethical concern of using Al in medical diagnosis and how it can be addressed.
Answer.Case Study: AI in Healthcare – X-ray Analysis
(a) Which type of neural network is most suitable for analyzing X-ray images?
Answer: b) CNN (Convolutional Neural Network) ✅
(b) Highlighting suspicious areas in an X-ray is an example of:
Answer: a) Image segmentation ✅
(c) Why is it important for doctors to review AI-generated results instead of relying solely on them?
Answer..Al tools, while powerful, are not infallible. They may produce false positives or negatives due to limitations in training data or unseen variations in medical conditions. Doctors bring human judgment, experience, and understanding of patient. history, which are essential to confirm diagnoses and avoid critical mistakes. Human oversight ensures patient safety and accountability in medical decision-making.
OR
Answer:
AI systems can make mistakes or produce false positives/negatives.
Doctors provide clinical context, judgment, and experience that AI cannot fully replicate.
Human review ensures patient safety and reduces the risk of misdiagnosis.
Combining AI and doctor expertise leads to faster yet reliable diagnoses.
(d) Discuss one ethical concern of using AI in medical diagnosis and how it can be addressed?
Answer.One major ethical concern is bias in Al models. If an Al system is trained on
unbalanced or non representative data (e.g. mostly from one ethnic group or age group), it may give inaccurate results for underrepresented populations How to address it:
*Use diverse and inclusive datasets during training
*Conduct regular audits to identify and correct bias.
*Maintain transparency in how Al makes decisions (explainable Al)
*Always involve human experts in the final decision-making process.
OR
Answer:
Ethical concern: Bias and fairness
AI may be biased if trained on data that is not diverse (e.g., only adults or certain ethnic groups).
This can cause misdiagnosis in underrepresented populations.
Solution:
Train AI models on diverse, high-quality datasets.
Regularly audit AI predictions to check for bias.
Keep doctors involved to verify AI results and ensure ethical patient care.
Q21.Sharmistha, a student of class X was exploring the Natural Language Processing domain. She got stuck while performing the text normalisation. Help her to normalise the text on the segmented sentences given below:
Document 1: Akash and Ajay are best friends.
Document 2: Akash likes to play football but Ajay prefers to play online games.
SHORT METHOD
Answer:-1. Tokenisation: Akash, and, Ajay, are, best, friends, Akash, likes, to, play, football, but Ajay, prefers, to, play, online, games.
2. Removal of stopwords: Akash, Ajay, best, friends, Akash, likes, play, football, Ajay,prefers, play, online, games
3. Converting text to a common case: akash, ajay, best, friends, akash, likes, play, football, ajay, prefers, play, online, games.
4. Stemming/Lemmatisation: akash, ajay, best, friend akash, like, play, football, ajay ,prefer, play, online, game.
LONG METHOD
Answer.
Given Sentences
Document 1:
Akash and Ajay are best friends.
Document 2:
Akash likes to play football but Ajay prefers to play online games.
Step 1: Sentence Segmentation
(Already given, isliye skip kar sakte hain)
Sentence 1: Akash and Ajay are best friends
Sentence 2: Akash likes to play football but Ajay prefers to play online games
Step 2: Tokenization
(Sentence ko words me todna)
Document 1 Tokens:
Akash | and | Ajay | are | best | friends
Document 2 Tokens:
Akash | likes | to | play | football | but | Ajay | prefers | to | play | online | games
Step 3: Lowercasing
(Sabhi words ko small letters me convert karna)
Document 1:
akash, and, ajay, are, best, friends
Document 2:
akash, likes, to, play, football, but, ajay, prefers, to, play, online, games
Step 4: Remove Punctuation
(Full stop, comma jaise symbols hata dena)
Document 1: . removed
Document 2: No punctuation
Step 5: Stop Words Removal
(Common words jaise: and, are, to, but)
Stop Words Removed
Document 1:
akash, ajay, best, friends
Document 2:
akash, likes, play, football, ajay, prefers, play, online, games
Step 6: Stemming / Lemmatization
(Word ko uske root form me lana)
likes → like
friends → friend
prefers → prefer
games → game
Final Normalised Text
Document 1:
akash, ajay, best, friend
Document 2:
akash, like, play, football, ajay, prefer, play, online, game