🔵 CHAPTER 1: UNDERSTANDING ARTIFICIAL INTELLIGENCE

👉 What is Artificial Intelligence?

The term Artificial Intelligence (AI) was given by John McCarthy in 1956.
According to him:
“AI is the science and engineering of making intelligent machines.”

📌 Simple meaning:

AI is a technology that allows machines to:

Think
Learn
Analyse
Make decisions
similar to human beings.

Examples around you:

Google Assistant answering questions
Amazon recommending products
Cars warning about obstacles
Face recognition in phones

🔵 Why is AI so popular today?

Earlier computers were slow, expensive, and stored small amounts of data.
But today we have:

✔ Ultra-fast processors

✔ Very large storage (Exabytes)

✔ High-speed data transfer (5G/6G, fibre optics)

✔ Ability to collect and store huge data

These advancements allow machines to learn effectively.

🔵 AI IN DAILY LIFE – DETAILED APPLICATIONS

AI is used in almost every industry.

1️⃣ Education & Training

AI is revolutionising teaching-learning.
Uses include:

Predicting student performance
Smart assessments
Identifying weak students
Robot-assisted teaching
Adaptive learning content (changes as per learner level)
Voice-enabled learning using NLP

2️⃣ E-Commerce (Amazon, Flipkart, Walmart)

AI uses huge customer data to:

Recommend products
Analyse user interests
Detect fraud orders
Track deliveries
Manage advertisements
Improve customer service

Example:
When you see shoes online, the site shows similar products → This is AI.

3️⃣ Social Media & Entertainment

AI is used for:

Automatic subtitles
Language translation
Editing videos
Detecting hate speech
Tracking content trends

4️⃣ Smart Emails

Gmail uses AI to:

Sort emails into categories (Primary, Social, Promotions)
Block spam
Suggest quick replies
Auto-complete sentences (“Smart Compose”)

5️⃣ Smart Chatbots & Voice Assistants

Examples: Alexa, Siri, Google Assistant
They use NLP to:

Understand your voice
Convert it into machine-readable data
Produce human-like responses

6️⃣ Futuristic AI

AI Nanobots for detecting hidden diseases
Intelligent prosthetics for disabled persons
Self-driving cars
Fraud detection systems

🔵 CHAPTER 2: DOMAINS OF ARTIFICIAL INTELLIGENCE

AI works mainly in three domains:

1️⃣ DATA DOMAIN (Brain of AI)

Data is the foundation of AI.
More data → Better predictions.

Example:

To predict sales of a mobile phone → AI uses previous years’ sale data, reviews, ratings, returns, etc.

Data helps AI identify patterns, trends, and relationships.

2️⃣ COMPUTER VISION (CV)

CV helps machines understand images & videos.

Uses:

Face recognition
CCTV surveillance
Barcode/QR code scanning
Photo-based search

How CV works:

Machine is trained with thousands of images → learns patterns → detects them in future.

3️⃣ NATURAL LANGUAGE PROCESSING (NLP)

NLP enables computers to understand human language (speech or text).

It involves:

Natural Language Understanding (NLU)
→ Understanding meaning
Natural Language Generation (NLG)
→ Machine generates human-like speech/text

Uses:

Google Translate
IVR systems
Chatbots
Voice typing

🔵 CHAPTER 3: AI PROJECT CYCLE (VERY IMPORTANT)

There are 5 major stages of the AI project cycle.

1️⃣ Problem Scoping

Identify and define the problem clearly.

Use the “4W Framework”:

Who → Stakeholders
What → Problem
Where → Context
Why → Purpose

Example (Library Problem from PDF):

Students take too long to find books → Create AI-based recommendation system.

2️⃣ Data Acquisition

Collecting relevant, useful, and quality data.

Features of good data:

✔ Relevant
✔ Accurate
✔ Up-to-date
✔ Large Volume
✔ Rich (variety)

Sources of Data:

Databases
Webpages
Sensors
CCTV
Forms & surveys
Web scraping

Methods to acquire data:

Export from databases
Scanning using OCR
Direct live feed (CCTV, sensors)
API

3️⃣ Data Exploration

Cleaning, organising, and understanding data.

Activities include:

Handling missing values
Removing duplicates
Formatting data
Feature engineering (extracting useful info)
Data visualisation (graphs)

Charts Used:

Bar chart (comparison)
Line chart (trends)
Pie chart (distribution)
Histogram (ranges)

4️⃣ Modelling

Building/choosing a model to make predictions.

Two approaches:

A. Rule-based Approach

Uses IF–ELSE statements
Does not learn
Limited intelligence

B. Learning-based Approach (ML & DL)

Learns from data
Handles complex tasks

Decision Tree (Important Model)

Structure contains:

Root Node (starting question)
Decision Node (yes/no branches)
Leaf Node (final result)

5️⃣ Evaluation

Evaluates how well the model performs using Confusion Matrix.

Confusion Matrix Terms

✔ True Positive (TP) – Predicted Yes & Actually Yes

✔ True Negative (TN) – Predicted No & Actually No

✔ False Positive (FP) – Predicted Yes but Actually No

✔ False Negative (FN) – Predicted No but Actually Yes

After successful evaluation → Model can be deployed as:

Software application
Web service
Mobile app
Robotics system

🔵 CHAPTER 4: DATA (TYPES, FORMATS, FEATURES)

👉 What is Data?

Data = raw facts that have no meaning unless arranged.
Example: “Raj, 16, 9, A”
Meaningful only when arranged → Raj is 16 years old, Class 9A.

Data Types:

Character (A, #, @)
String/Text (“India”)
Numbers (10, 20, -1)

Complex Data Types:

Audio
Video
Images (JPG/PNG)
3D models

Data Formats:

Dates
Decimals
Uppercase/lowercase

🔵 CHAPTER 5: EVALUATION & MODEL DEPLOYMENT

After modelling, testing is done using testing data.

If the model performs well → deploy it.

Forms of Deployment:

Mobile app
Full software
Web service
Part of robotics system

🔵 CHAPTER 6: AI ETHICS

Ethics = moral principles.

Why ethical AI?

AI is powerful → can cause harm if misused.

Challenges:

1. Accountability:

Who is responsible if AI harms someone?

2. Bias & Transparency:

AI may show discrimination if training data was biased.

3. Pre-existing Bias:

Data collected from society already contains bias.

4. Technical Bias:

Errors due to incorrect algorithms or insufficient data.

5. Emergent Bias:

Bias occurs during real usage (e.g., app difficult for elderly).

6. Security & Privacy:

AI systems may be hacked; data may be stolen.