Scribd
Upload
6 views

Notes unit-2

The document outlines key concepts in AI, including differences between Rule-Based and Learning-Based models, and explains various types of machine learning such as Supervised, Unsupervised, and Reinforcement Learning with examples. It also discusses the structure and function of Neural Networks, differentiates between Classification and Regression models, and compares Artificial Neural Networks (ANN) with Convolutional Neural Networks (CNN). Additionally, it highlights the distinctions between Machine Learning and Deep Learning.

Uploaded by

aaditri.iv
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
6 views

Notes unit-2

The document outlines key concepts in AI, including differences between Rule-Based and Learning-Based models, and explains various types of machine learning such as Supervised, Unsupervised, and Reinforcement Learning with examples. It also discusses the structure and function of Neural Networks, differentiates between Classification and Regression models, and compares Artificial Neural Networks (ANN) with Convolutional Neural Networks (CNN). Additionally, it highlights the distinctions between Machine Learning and Deep Learning.

Uploaded by

aaditri.iv
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 4

Q1. What is the difference between Rule-Based and Learning-Based AI models?

Ans:
In Rule-Based AI models, the system follows predefined rules set by developers and does not
learn from data. In contrast, Learning-Based models learn patterns from data and improve over
time by adapting to new information.

Q2. Define Supervised Learning with an example.


Ans:
Supervised Learning is a type of machine learning where the model is trained using labeled
data. For example, a model trained to recognize fruits by using images labeled as “apple” or
“banana”.

Q3. How is Unsupervised Learning different from Supervised Learning?


Ans:
Unsupervised Learning works with unlabeled data and identifies hidden patterns, whereas
Supervised Learning uses labeled data to train the model to predict known outputs.

Q4. Give an example of Reinforcement Learning.


Ans:
An AI playing a game and improving its performance by receiving rewards for correct moves and
penalties for mistakes is an example of Reinforcement Learning.

Q5. What are Features and Labels in a dataset?


Ans:
Features are input variables used to predict outcomes (e.g., weight, color). Labels are the target
output values (e.g., fruit name) assigned to the data.

Q6. What is the use of a training dataset?


Ans:
A training dataset is used to teach the AI model by providing examples, so the model can learn
patterns and make accurate predictions.

Q7. What are the two types of Supervised Learning models?


Ans:
The two types are:

1. Classification Model – predicts discrete labels (e.g., spam or not).

2. Regression Model – predicts continuous values (e.g., house price).


Q8. Explain Clustering with an example.
Ans:
Clustering groups similar data points. For example, a music app clustering songs by tempo and
intensity to suggest playlists based on user preferences.

Q9. What is the role of Neural Networks in Deep Learning?


Ans:
Neural Networks mimic the human brain's structure to process large datasets automatically
and extract features without human intervention, useful in tasks like image recognition.

Q10. Name the layers of a Neural Network and state their function.
Ans:

• Input Layer: Accepts data.

• Hidden Layers: Perform computations using weights and biases.

• Output Layer: Delivers the final prediction or result.

Q1. Explain the differences between Supervised, Unsupervised, and Reinforcement


Learning with one example each.

Learning Type Definition Data Type Example


Spam email
Supervised Model is trained using labeled data to
Labeled detection using
Learning predict outcomes.
tagged emails.
Grouping users
Unsupervised Model identifies hidden patterns in
Unlabeled based on browsing
Learning unlabeled data.
history.
Model learns by interacting with the AI learning to play a
Reinforcement Feedback-
environment through game through trial
Learning based
rewards/punishments. and error.

Q2. What is a Neural Network? Explain its three main layers and their functions.

Answer:
A Neural Network is a computing system inspired by the human brain that processes
information through interconnected nodes (neurons). It automatically learns patterns from
large datasets.

Layer Function
Input Layer Receives raw data and sends it to the hidden layers.
Hidden Layers Process the data using weights, biases, and activation functions.
Output Layer Displays the final result or prediction to the user.
Q3. Compare Rule-Based and Learning-Based AI Models. State one real-world
example for each.

Feature Rule-Based AI Model Learning-Based AI Model


Approach Follows predefined rules Learns from data and adapts
Data
Does not require large data Requires a large dataset for training
Requirement
Static – cannot improve with Dynamic – improves with feedback and
Adaptability
new data new data
Spam filter that improves with user
Example Website chatbot for FAQs
feedback

Q4. Differentiate between Classification and Regression models with suitable examples.

Model
Classification Regression
Type
Output
Categorical (Discrete) Continuous value
Type
Data Used Labeled data Labeled data
Predicting if an email is spam or Predicting house price based on size and
Example
not location
Use Case Sorting items into classes Estimating numerical outcomes

Q5. What is Clustering in Machine Learning? How is it different from Classification?


Support your answer with an example.

Feature Clustering (Unsupervised) Classification (Supervised)


Data Unlabeled Labeled
Process Groups data by similarity Assigns data to predefined classes
Learning Type No prior knowledge of output Learns from labeled data
Example Grouping songs by tempo and mood Tagging emails as spam or not spam

Q 6 Difference between ANN and CNN

Convolutional Neural Network


Feature Artificial Neural Network (ANN)
(CNN)

Full Form Artificial Neural Network Convolutional Neural Network

Made up of fully connected layers Includes convolutional layers, pooling


Structure
(input, hidden, output) layers, and fully connected layers
Convolutional Neural Network
Feature Artificial Neural Network (ANN)
(CNN)

Structured/tabular data (e.g., stock Image, video, or visual data


Best Suited For
prediction, text classification) processing (e.g., object detection)

Feature Manual – features must be given as Automatic – extracts features from


Extraction input raw images

Computation Generally less computationally More computationally intensive due to


Cost expensive matrix operations

Works well with spatial (2D/3D) data


Data Handling Works well with flat (1D) data
like images

No – Each neuron has its own Yes – Filters share weights across the
Weight Sharing
weights input

Example Face recognition, self-driving cars,


Fraud detection, spam classification
Applications medical image analysis

Q7 Difference between Machine Learning and Deep Learning

Feature Machine Learning (ML) Deep Learning (DL)

A subset of AI that enables A subset of ML that uses neural


Definition
machines to learn from data networks to process large data sets

Works well with small to medium Requires large amounts of data for
Data Dependency
datasets training

Manual – features must be


Feature Extraction Automatic – extracts features on its own
selected by humans

Training Time Relatively short Requires longer training time

Algorithm Uses simpler algorithms like Uses complex neural networks with
Complexity decision trees, SVM, etc. multiple layers

Easier to interpret and explain Often seen as a "black box" – harder to


Interpretability
results interpret

Hardware
Can run on standard CPUs Needs high-performance GPUs
Requirement

Example Email filtering, loan approval Image recognition, speech-to-text,


Applications prediction autonomous vehicles

You might also like