Unsupervised Learning

Unsupervised machine learning involves inputting unlabeled data and allowing the algorithm to find hidden patterns in the data without human intervention. Common techniques include clustering, which groups similar data points together, and dimensionality reduction, which simplifies complex data. Clustering algorithms like k-means partitioning and hierarchical clustering are often used to analyze customer data for marketing or group online users based on shared interests without supervision. Unsupervised learning has applications in fields such as bioinformatics, biology, data mining, and computer vision.

Uploaded by

Umar Khan

We take content rights seriously. If you suspect this is your content, claim it here.

Available Formats

Download as PPTX, PDF, TXT or read online on Scribd

0% found this document useful (0 votes)

34 views

Unsupervised Learning

Uploaded by

Umar Khan

We take content rights seriously. If you suspect this is your content, claim it here.

Available Formats

Download as PPTX, PDF, TXT or read online on Scribd

You are on page 1/ 14

Unsupervised

Machine Learning
Introduction
• A type of Machine Learning
• Input Features without labeled target attribute
• Draw inferences from datasets
• Explore the data to find some intrinsic features
Examples
• Grouping of people that have similar sizes so make only 3

categories of shirts for the customers like “small”, “medium”

and “large”.

• Divide the group of customers according to their similarities in

buying behavior for the targeted marketing.

• Grouping of communities on social media based on their

interest.
Clustering
• Most common method of Unsupervised learning
• It is used to analyze data features to find hidden patterns by
forming groups or clusters in data.
• The clusters are modeled using a measure of similarity which
is defined upon metrics such as Euclidean or probabilistic
distance.
Clustering Algorithms
• Partitioning clustering: partitions data into k distinct clusters
based on distance to the centroid of a cluster.
• Hierarchical clustering: builds a multilevel hierarchy of clusters
by creating a cluster tree.
• Model Based: Hypothesize a model for each cluster and find
best fit of the models to data.
• Density Based: Guided by connectivity and density functions.
• Gaussian mixture models: models clusters as a mixture of
multivariate normal density components.
• Hidden Markov models: uses observed data to recover the
sequence of states.
Partitioning Clustering
• Using K means Algorithm.
• Partition the data into k clusters.
• Each data point associate itself with the nearest mean (cluster
centers or cluster centroid), serving as a prototype of the
cluster.
• Centroids are updating with the addition and subtraction of
data points in clusters.
• Recursively doing the process, gives the optimized clusters.
• Stops when there is a minimum or no change in centroids
values.
Contd.
Hierarchal Clustering
• This algorithm produces a nested sequence of clusters.

• Algorithm compute a distance matrix between different data

points.
• Based on low value between data points, the cluster is formed
between them.
• Recursively doing the process, gives a hierarchal structure

• Each cluster is distinct from each other and the data in each
cluster are broadly similar to each other.
Contd.
Types of Hierarchal Clustering
• Agglomerative (bottom-up) Clustering:
It builds the dendogram from the bottom level and merges the
most similar or nearest pair of clusters.
Stop when all the clusters are merged into single root node.

• Divisive (top down) Clustering:

It starts with all data points in one root node.
Splits the root node into a set of child clusters. Each child node is
recursively divided further into clusters.
Stops when only singleton clusters of individuals data points
remain, i.e. each cluster with only a single point.
Contd.
Applications