Scribd
Upload
22 views

Unsupervised Learning

Uploaded by

anupriyasoundararaj
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
22 views

Unsupervised Learning

Uploaded by

anupriyasoundararaj
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 21

Unsupervised Learning

in Machine Learning
GRADE - 11
What do you
already know about
unsupervised
learning?
Imagine sorting thousands of
mystery items by hand with no
labels. How could a computer do
this faster and more effectively?
Comprehend the principles of
Unsupervised Learning.

WALT:
Understand real-world
applications and types of
Unsupervised Learning.
Unsupervised Learning

Clustering
Keywords:
Dimensionality Reduction

Data Patterns

Anomalies
Explain I can explain what Unsupervised Learning is.

I can list key types and algorithms in


List Unsupervised Learning.
Success
Discuss
I can discuss real-world applications for Criteria
Unsupervised Learning.

I can contrast Unsupervised and Supervised


Contrast Learning approaches.
Introduction to
A type of machine learning that
Unsupervised
finds hidden patterns or
Learning: intrinsic structures in data
without any labels.

To analyze and group data by


discovering relationships
without guidance.
Model is trained on
data without labeled
How outputs.
Unsupervised
Learning Works? It finds hidden
patterns, clusters, or
structures in data.
Think – Pair – Share !

Discuss with a partner why finding


patterns without labels could be useful.
Share your ideas with the class.
Types of
Unsupervised Clustering: Grouping data points
based on similarity (e.g.,
Learning
customer segmentation).

Dimensionality Reduction:
Simplifying data while
preserving important structure
(e.g., PCA for reducing features).
Examples of Brainstorm with a partner a real-
life example of clustering and
Unsupervised dimensionality reduction and
Learning explain why it fits.
Key Algorithms in
Unsupervised Learning
Which algorithm do you think
might work best for customer
Clustering segmentation, and why?

Dimensionality
Reduction
Clustering
Types of Clustering:
• K-Means Clustering: Partitions
Clustering groups similar data into a specified number of
data points together. clusters.
• Hierarchical Clustering: Creates
a tree-like structure of clusters.

Imagine you have a collection of photos on your phone. How could clustering
help you organize them automatically?
Reduces data complexity by decreasing the number of features.
Common Techniques:
▪ Principal Component Analysis (PCA):
Reduces dimensions by emphasizing variance.
▪ t-Distributed Stochastic Neighbor Embedding (t-SNE):
Dimensionality Maps data points to lower dimensions for easier visualization.

Reduction
Imagine you’re designing a mobile app
that needs to use fewer features to save
memory. How would you decide which
features are most important?
Real-Life Applications of Unsupervised
Learning
Anomaly Detection:
Market Segmentation: Recommendation
Identifying unusual
Grouping customers Systems: Clustering
patterns in data for
based on purchasing similar items for better
security or
behavior. recommendations.
maintenance.

Which example do you find the most interesting?


Group Discussion and Sharing
Small groups, brainstorm two more uses of
unsupervised learning in everyday life. Share your
ideas with the class.

What questions do you


What new insights have still have about clustering
you gained from hearing or dimensionality
others’ examples? reduction?
Problem Statement:
Imagine you’re a data scientist grouping customers of
an online bookstore based on their browsing patterns.
Steps:
Example Problem 1) Choose the number of clusters.
– 2) Calculate centroids for each cluster.
Applying K-Means 3) Group data points based on similarity to centroids.
Clustering
Sketch how you might group
customers based on age and
book categories.
Evaluating Unsupervised Learning Models

Silhouette Score: Visualization: Plot


Measures how data to visually
well-defined the inspect cluster
clusters are. separation.

Why might it be challenging to assess the performance of


an unsupervised model compared to a supervised one?
All Levels: Write a concise definition of unsupervised learning and list
two principles with examples.

Level 1: Draw a mind map explaining the types of unsupervised learning


(clustering and association) with brief descriptions and one example
each.

Level 2: Use art to creatively represent clustering. For example, draw a


Task Time ! visual of grouped data points (e.g., colorful balloons grouped by size and
color) and label the clusters with meaningful names.

Level 3: Prepare a 3-slide presentation explaining how clustering is used


in an industry of your choice (e.g., healthcare, retail, or education).
Include a real-world example, a simple dataset visualization, and a
conclusion slide.
Explain I can explain what Unsupervised Learning is.

List
I can list key types and algorithms in
Unsupervised Learning. Revisiting
Success
Discuss
I can discuss real-world applications for
Unsupervised Learning. Criteria
I can contrast Unsupervised and Supervised
Contrast Learning approaches.
Hold up 1 finger for True
and 2 fingers for False.
True or False!
1. In unsupervised learning, the data used to train the model has labels.
2. Clustering is used to group similar data points without needing predefined
categories.
3. Dimensionality reduction increases the number of features in a dataset.
4. An example of unsupervised learning is identifying clusters of customers based on
shopping patterns.
5. K-Means clustering requires you to choose the number of clusters before running
the algorithm.
6. In dimensionality reduction, Principal Component Analysis (PCA) is commonly used
to reduce features by focusing on variance.
7. Unsupervised learning can’t be used for anomaly detection.

You might also like