1.

Introduction to Deep Learning

Deep Learning is a part of Artificial Intelligence (AI) that teaches computers to learn from data
in a way similar to the human brain. It uses special models called neural networks that can
automatically find patterns in large amounts of data. Deep learning is widely used in image
recognition, speech recognition, medical diagnosis, and self-driving cars.

2. Relationship Between AI, Machine Learning, and Deep Learning

 Artificial Intelligence (AI): The broad field of making machines intelligent.

 Machine Learning (ML): A subset of AI where machines learn from data instead of being
programmed step by step.

 Deep Learning (DL): A subset of Machine Learning that uses many layers of neural
networks to learn complex patterns.

In simple words, Deep Learning is a more advanced form of Machine Learning.

AI, Machine Learning, and Deep Learning

1. Artificial Intelligence (AI)

Artificial Intelligence is the main and broad field. It focuses on creating machines or systems
that can think, reason, and make decisions like humans. AI does not always need learning from
data; it can also work using fixed rules and logic.

Key Points:

 Broadest concept

 Can work with rules or logic

 Goal is to make machines intelligent

Example:

A chess-playing computer that follows pre-defined rules.

2. Machine Learning (ML)

Machine Learning is a subset of AI. It allows machines to learn from data and improve
performance without being programmed step by step. ML systems find patterns in data and
make predictions.

Key Points:

 Part of AI

 Learns from data

 Needs human effort for feature selection

Example:

An email spam filter that learns from previous emails.

3. Deep Learning (DL)

Deep Learning is a subset of Machine Learning. It uses deep neural networks with many
hidden layers to automatically learn complex patterns. It requires large amounts of data and
powerful computers.

Key Points:

 Part of Machine Learning

 Uses multi-layer neural networks

 Automatically extracts features

Example:

Face recognition systems in mobile phones.

Summary Table

Aspect Artificial Intelligence (AI) Machine Learning (ML) Deep Learning (DL)

Scope Broadest field Subset of AI Subset of ML

Learning May or may not learn Learns from data Learns from large
data

Human Involvement High Medium Low

Complexity Low to high Medium High

Data Requirement Low Medium Very high

Example Rule-based chatbot Spam detection Image recognition

3. What Are Neural Networks?

A Neural Network is a computer model inspired by the human brain. It is made of small units
called neurons that work together to process information.

Basic Structure of a Neural Network

 Input Layer: Takes the raw data (such as an image or text).

 Hidden Layers: Process the data and learn patterns.

 Output Layer: Gives the final result (such as a label or prediction).

When a network has many hidden layers, it is called a deep neural network, and this is the
basis of deep learning.

4. How Deep Learning Works

Deep learning works by learning from examples:

1. Data is given to the neural network.

2. The network makes a prediction.

3. The prediction is compared with the correct answer.

4. Errors are calculated.

5. The network adjusts its weights to reduce errors.

This process is repeated many times until the model becomes accurate. This learning method is
called training.

5. Key Components of Deep Learning

5.1 Data

Deep learning needs a large amount of data to work well. More data helps the model learn
better patterns.

5.2 Weights and Bias

 Weights: Decide how important an input is.

 Bias: Helps the model fit the data better.

5.3 Activation Functions

Activation functions decide whether a neuron should be activated or not.

Common activation functions include:

 ReLU

 Sigmoid

 Softmax

6. Types of Deep Learning Models

6.1 Artificial Neural Networks (ANN)

Used for basic prediction and classification problems.

6.2 Convolutional Neural Networks (CNN)

Specially designed for image and video processing.

Example: Face recognition, medical image analysis.

6.3 Recurrent Neural Networks (RNN)

Used for sequence data like text and speech.

Example: Language translation, speech recognition.

7. Applications of Deep Learning

 Image Recognition: Identifying faces or objects in images.

 Speech Recognition: Voice assistants like Siri and Google Assistant.

  Healthcare: Detecting diseases from medical images.

 Self-Driving Cars: Detecting roads, pedestrians, and traffic signs.

 Recommendation Systems: Netflix and YouTube recommendations.

8. Advantages of Deep Learning

 Can handle very large and complex data.

 Automatically learns features without manual effort.

 Provides high accuracy for complex tasks.

9. Limitations of Deep Learning

 Requires a large amount of data.

 Needs powerful hardware (GPUs).

 Training takes a lot of time.

 Models are often hard to understand (black box).

Expert Systems and Intelligent Agents in Artificial Intelligence

Expert Systems

Expert systems are computer programs that act like human experts in a specific field. They use
stored knowledge and if–then rules to solve problems and give decisions, just like a human
specialist would do.

These systems have two main parts:

 Knowledge Base: Stores expert information and rules.

 Inference Engine: Uses the rules to make decisions.

Example:

A medical diagnosis system that suggests diseases based on symptoms provided by the patient.

Intelligent Agents
Intelligent agents are AI systems that can observe their environment, make decisions, and take
actions to achieve a specific goal. They continuously interact with their surroundings and
improve their performance.

Main features:

 Sense the environment

 Decide what action to take

 Act to achieve goals

Example:

Virtual assistants like Google Assistant or Siri, which listen to user commands and perform
tasks.

Below is a detailed yet easy-to-understand answer on Machine Learning, written with clear
headings, simple wording, and examples, suitable for exams and assignments.

Machine Learning (ML)

1. Introduction to Machine Learning

Machine Learning is an important branch of Artificial Intelligence (AI) that enables computers
to learn from data and improve their performance without being explicitly programmed.
Instead of following fixed rules, ML systems identify patterns in data and make predictions or
decisions based on experience.

2. How Machine Learning Works

Machine learning works in the following steps:

1. Data Collection: Large amounts of data are gathered.

2. Training: The machine analyzes data and learns patterns.

3. Model Creation: A model is built using learning algorithms.

4. Testing: The model is tested with new data.

 5. Prediction: The trained model makes decisions or predictions.

With more data and practice, the system becomes more accurate.

3. Types of Machine Learning

3.1 Supervised Learning

In supervised learning, the machine is trained using labeled data (data with correct answers).

Examples:

 Email spam detection

 Student result prediction

Common algorithms: Linear Regression, Decision Trees

3.2 Unsupervised Learning

In unsupervised learning, the machine learns from unlabeled data and finds hidden patterns.

Examples:

 Customer grouping (clustering)

 Market segmentation

Common algorithms: K-means, Hierarchical Clustering

3.3 Reinforcement Learning

In reinforcement learning, the machine learns by trial and error. It receives rewards for correct
actions and penalties for wrong actions.

Examples:

 Game-playing AI

 Robot navigation

4. Common Machine Learning Algorithms

  Linear Regression

 Logistic Regression

 Decision Trees

 Support Vector Machines (SVM)

 K-Nearest Neighbors (KNN)

5. Applications of Machine Learning

 Healthcare: Disease prediction, medical imaging

 Finance: Fraud detection, credit scoring

 Education: Student performance analysis

 Business: Sales forecasting, customer behavior analysis

 Social Media: Content recommendations

6. Advantages of Machine Learning

 Improves accuracy over time

 Handles large datasets efficiently

 Automates decision-making

 Reduces human effort

7. Challenges of Machine Learning

 Requires large and good-quality data

 Risk of biased results

 High computational cost

 Models can be difficult to interpret

8. Difference Between Machine Learning and Traditional Programming

Traditional Programming Machine Learning

Fixed rules are written Rules are learned from

data

No learning Learns from experience

Less flexible More adaptive

Below is an easy but slightly detailed explanation of Ethical and Professional Issues in AI, with
simple examples. This is suitable for 5–8 mark exam answers.

Ethical and Professional Issues in Artificial Intelligence (AI)

1. Data Privacy

AI systems use a lot of personal data such as names, medical records, and phone numbers. If
this data is not stored and used properly, people’s privacy can be violated.

Example:

A healthcare AI system leaking patient medical records.

2. Bias and Unfair Decisions

If AI is trained on biased or incomplete data, it can make unfair decisions that favor one group
over another.

Example:

A hiring AI system selecting mostly male candidates.

3. Lack of Transparency

Many AI systems do not clearly explain how they make decisions. This makes it hard for users to
trust them.

Example:
An AI system rejecting a bank loan without explaining the reason.

4. Responsibility and Accountability

When AI makes a mistake, it is unclear who should be blamed—the programmer, the company,
or the AI system.

Example:

A self-driving car accident where no clear responsibility is defined.

5. Job Displacement

AI automation can replace human workers, causing job loss in some industries.

Example:

Machines replacing workers in factories.

6. Misuse of AI

AI can be used for harmful purposes if not controlled properly.

Example:

Deepfake videos used to spread false information.

7. Professional Responsibility

AI professionals must design systems that are safe, fair, and ethical.

Example:

Developers testing AI systems to avoid biased or harmful results.

Got it! I will now provide separate sets with slightly longer, detailed but easy-to-understand
answers for each group of questions. Each answer will be short paragraphs, suitable for exams.

Set 1
1. Explain the difference between Automation and Artificial Intelligence.

 Automation refers to the use of machines or software to perform repetitive tasks based
on pre-defined rules. It cannot make decisions or adapt to new situations.

 Artificial Intelligence (AI) allows machines to learn from data, make decisions, and adapt
to new situations without explicit programming.
Example: Automation → a factory conveyor belt; AI → a self-driving car that adapts to
traffic conditions.

Feature Automation Artificial Intelligence (AI)

Core Executes repetitive, rule-based tasks. Simulates human cognitive

Function functions.

Logic Follows "If-Then" rules (pre-programmed). Learns from data and adapts
(probabilistic).

Example A factory arm painting a car the same way A self-driving car navigating a new
every time. city.

Think of the difference like this:

 Automation is a Follower: It follows a fixed set of rules.

o Example: An alarm clock. You set it for 7:00 AM, and it rings every day at 7:00
AM. It doesn't "think"—it just obeys.

 AI is a Learner: It looks at data and makes its own decisions.

o Example: A "Smart" alarm clock that checks the traffic and wakes you up 20
minutes earlier because there’s a crash on your route. It "learned" that you
needed more time.

2. List any two major advancements in IT during the 2020s.

 Cloud Computing: Storing and accessing data and applications over the internet instead
of local servers.

 Artificial Intelligence & Machine Learning: Intelligent systems that analyze data and
automate tasks in various fields.
3. Difference between Machine Learning and Deep Learning.

 Machine Learning (ML): Uses algorithms to learn from data, often requires manual
selection of features.

 Deep, Learning (DL): Uses multi-layer neural networks to learn complex patterns
automatically, usually requires large datasets.

4. What are the types of AI? Give one example each.

 Narrow AI: Focused on one task. Example: Virtual assistants like Siri.

 General AI: Can perform multiple tasks like humans. Example: Not fully developed yet.

 Super AI: Smarter than humans. Example: Conceptual/future AI.

5. List one benefit and one risk of using AI in modern systems.

 Benefit: AI can automate tasks and improve efficiency.

 Risk: AI may cause job displacement or make biased decisions if trained on biased data.

Set 2

1. Why is Artificial Intelligence considered a core technology in modern IT ecosystems?

AI is core to modern IT because it enables automation, smart decision-making, and data

analysis. It helps organizations handle large data, improve efficiency, and develop intelligent
applications like chatbots, recommendation systems, and autonomous vehicles.

2. Distinguish between Narrow AI and General AI.

 Narrow AI: Designed for specific tasks. Cannot perform tasks outside its scope. Example:
Google Translate.

 General AI: Can perform multiple tasks like a human, with flexible thinking. Still under
development.

3. Describe the basic working pipeline of an AI system.

1. Data Collection: Gathering relevant data.

2. Data Preprocessing: Cleaning and organizing data for use.

3. Model Training: Using algorithms to learn patterns from data.

4. Testing: Evaluating the model’s performance on new data.

 5. Prediction/Decision: Applying the trained model to make decisions or predictions.

4. Give two real-world applications of AI and explain how they improve efficiency.

 Healthcare: AI helps detect diseases from medical images faster and more accurately
than humans, saving time and improving patient care.

 Business: AI chatbots respond to customer queries instantly, reducing human workload

and improving response time.

5. Explain one ethical issue related to Artificial Intelligence.

 Bias: AI systems can make unfair or discriminatory decisions if trained on biased data.
Example: Hiring systems that favor one gender or group over others.

Set 3

1. Define Artificial Intelligence. With example.

Artificial Intelligence is the ability of machines to perform tasks that normally require human
intelligence, such as learning, reasoning, and problem-solving.

Example: Self-driving cars that navigate and make decisions on roads.

2. What is the difference between ML and DL?

 ML: Learns from data using algorithms, usually needs manual feature selection.

 DL: Uses deep neural networks to automatically learn complex patterns from large data.

3. What is Generative AI? Example.

Generative AI is a type of AI that can create new content like images, text, or music, based on
patterns learned from data.

Example: ChatGPT generating text, DALL-E creating images from prompts.

4. Explain Agentic AI. With example.

Agentic AI refers to AI systems that perceive their environment, make decisions, and act
autonomously to achieve goals.

Example: A robot vacuum that maps a room and cleans automatically.

5. Mention one challenge of AI in recent era.

  Lack of transparency in decision-making. Many AI systems are “black boxes,” making it
hard to understand or trust their outputs.

Set 4

1. Name any two global IT trends of 2025.

 Artificial Intelligence and Machine Learning: Increasingly used in business, healthcare,

and education.

 Metaverse and Extended Reality (XR): Virtual environments for gaming, collaboration,
and education.

2. What is the role of IT in healthcare?

IT helps manage patient records, provides telemedicine services, analyzes health data, and
supports AI-driven diagnosis and treatment, improving efficiency and accuracy.

3. Define “digital ecosystem”. Example.

A digital ecosystem is a network of connected digital services, platforms, and applications that
work together.

Example: Google Workspace connects Gmail, Drive, Docs, and Calendar for seamless work.

4. Explain one IT career path.

 Data Scientist: Collects, analyzes, and interprets data using AI and ML to help
organizations make informed decisions.

5. One challenge faced by global IT systems.

 Cybersecurity threats: Hackers and malware can compromise data privacy and system
integrity.

Here’s an easy, clear explanation of the types of AI, including Narrow, General, Super, Agentic,
and Generative AI, with examples:

Types of Artificial Intelligence (AI)

1. Narrow AI (Weak AI)

 Performs specific tasks only.

 Cannot perform tasks outside its designed function.

  Most AI today is Narrow AI.
Example:

 Siri or Alexa answering questions

 Google Translate translating text

2. General AI (Strong AI)

 Can perform multiple human-like tasks.

 Can learn, reason, and adapt like a human.

 Still under research; not fully developed.

Example:

 An AI that could drive, teach, and answer questions like a human (conceptual).

3. Super AI

 Smarter than humans in every task.

 Can make independent decisions and outperform humans.

 Currently theoretical and does not exist yet.

Example:

 Advanced AI in science fiction that controls cities or makes inventions beyond human
capability.

4. Agentic AI

 Can sense its environment, make decisions, and act autonomously to achieve goals.

 Focuses on independent action and decision-making.

Example:

 Robot vacuum cleaners that map rooms and clean automatically

 Autonomous drones performing tasks without human control

5. Generative AI
  Can create new content such as text, images, videos, or music.

 Learns patterns from existing data and generates something new.

Example:

 ChatGPT generating text

 DALL-E creating images from prompts

 AI music generators composing songs

Summary Table:

Type of AI Key Feature Example

Narrow AI Performs specific task only Siri, Alexa

General AI Can perform human-like tasks Conceptual AI

Super AI Smarter than humans Sci-fi AI concepts

Agentic AI Acts autonomously in Robot vacuum, autonomous

environment drone

Generative AI Creates new content ChatGPT, DALL-E