UNSUPERVISED LEARNING

NIVEDITHA. GH
VI Semester
MACHINE LEARNING
 Unsupervised
learningUNSUPERVISED LEARNING
 Unsupervised learning is a type of machine learning where the algorithm learns patterns and
structures from input data without explicit supervision or labeled responses.
 In unsupervised learning, the algorithm is given a dataset consisting of features but not the
corresponding target labels. The goal is to explore the inherent structure of the data, identify
patterns, and extract meaningful insights.
 Unsupervised Learning is the training of a machine using information that is neither classified nor
labeled , which allows the algorithm to act on that information without guidance.
TYPES OF UNSUPERVISED LEARNING
CLUSTERING: ASSOCIATION:

 Clustering is a method of  An association rule is an

grouping the objects into unsupervised learning method
clusters such that which is used for finding the
objects with most relationships between variables
similarities remains into in the large database.
a group and has less or
no similarities with the  It determines the set of items
objects of another that occurs together in the
group. dataset. Association rule makes
marketing strategy more
effective.
 Cluster analysis finds the
commonalities between  Such as people who buy X item
the data objects and (suppose a bread) are also tend
categorizes them as per to purchase Y (Butter/Jam)
the presence and item. A typical example of
absence of those Association rule is Market
commonalities. Basket Analysis.
 SUPERVISED LEARNING VS UNSUPERVISED LEARNING
Supervised learning algorithms are trained using Unsupervised learning algorithms are trained using
labeled data unlabeled data.
Supervised learning model predicts the output. Unsupervised learning model finds the hidden patterns
in data.
The goal of supervised learning is to train the model so The goal of unsupervised learning is to find the hidden
that it can predict the output when it is given new data. patterns and useful insights from the unknown dataset.
Supervised learning can be categorized in Classification Unsupervised Learning can be classified in Clustering
and Regression problems. and Associations problems.
Supervised learning model produces an accurate result. Unsupervised learning model may give less accurate
result as compared to supervised learning.
It includes various algorithms such as Linear Regression, It includes various algorithms such as Clustering, KNN,
Logistic Regression, Support Vector Machine, Multiclass and Apriori algorithm.
Classification, Decision tree, Bayesian Logic, etc.
Supervised learning needs supervision to train the Unsupervised learning does not need any supervision
model. to train the model.
Supervised learning is not close to true Artificial Unsupervised learning is more close to the true
intelligence as in this, we first train the model for each Artificial Intelligence as it learns similarly as a child
data, and then only it can predict the correct output. learns daily routine things by his experiences.
 GENOME ANALYSIS: These Algorithms can analyze genetic data to uncover
patterns and links for genetic research.

IMAGE AND TEXT CLUSTERING: It can automatically group similar images and
texts , aiding in tasks like image organization, document clustering or content
recommendation.
APPLICATIONS CUSTOMER SEGMENTATION: Customers can be grouped based on their purchase
OF behavior, allowing organizations to customize marketing efforts.

UUNSUPERVISE SEMANTIC CLUSTERING: It organizes all the responses with the same meaning
into clusters to ensure that customer quickly and easily gets the information.
D LEARNING
ANOMALY DETECTION: Unsupervised learning is used to identify data
points, events, and/or observations that deviate from a dataset's normal
behavior.

MARKET BASKET ANALYSIS: Past purchase behavior coupled with

unsupervised learning can be used to help businesses discover data trends
that they could use to develop effective cross-selling strategies.
 CLUSTERING
o Clustering in unsupervised machine learning, is the process
of grouping unlabeled data into clusters, based on their
similarities.

o Broadly this technique is applied to group data, based on

different patterns, such as similarities or differences, our
machine model finds.

o Clustering aims at forming groups of homogeneous data

points from a heterogeneous dataset. It evaluates the
similarity based on a metric like Euclidean distance, Cosine
similarity, Manhattan distance, etc. and then group the
points with highest similarity score together.

PARTIONING BASED CLUSTERING: These methods divide the data set into distinct groups. They aim to create partitions
such that objects within each partition are as similar as possible.
 TYPES OF CLUSTERING
•K-means Clustering: Partitions data into k clusters based on centroid minimization.
•K-medoids Clustering: Similar to k-means but uses medoids (most centrally located points) instead of centroids.

HIERARCHICAL CLUSTERING: These methods create a hierarchy of clusters that can be represented in a tree structure
(dendrogram).
•Agglomerative (Bottom-Up): Starts with each object in its own cluster and merges the closest pairs of clusters iteratively.
•Divisive (Top-Down): Starts with all objects in one cluster and splits the cluster iteratively.

DENSITY-BASED CLUSTERING: These methods find clusters based on the density of points in the data space.
•DBSCAN (Density-Based Spatial Clustering of Applications with Noise): Finds clusters based on the density of data points,
capable of finding arbitrary-shaped clusters and dealing with noise.
•OPTICS (Ordering Points to Identify the Clustering Structure): An extension of DBSCAN that works better with varied
densities.

GRID BASED CLUSTERING: These methods involve dividing the data space into a finite number of cells that form a grid
structure.
•STING (Statistical Information Grid): Uses a multi-resolution grid data structure for efficient clustering.
•CLIQUE (Clustering In QUEst): Identifies dense regions in subspaces of high-dimensional data and merges them to form
clusters.
 APPLICATIONS
MARKETING: Clustering can be used for marketing reasons to characterize and discover
client segments.

BIOLOGY: It can be used for classification among different species of plants and animals.

IMAGE PROCESSING: It can be used to group similar images together, classify images based
on content and identify patterns in image data.

FRAUD DETECTION: Clustering is used to identify suspicious patterns or anomalies in

financial transactions, thereby helping in fraud detection.

CYBERSECURITY: It is used to group similar patterns of network traffic or system behavior

,which can help in detecting and preventing cyber attacks.

CLIMATE ANALYSIS: It is used for grouping comparable patterns of climate data, such as
temperature, precipitation and wind inorder to better understand climate change and its
influence on environment.

MEDICAL DIAGNOSIS: It is used to group patients with similar symptoms or diseases, Which helps in making accurate diagnoses and
identifying effective treatments.
 K-MEANS CLUSTERING
FINANCE: Clustering is technique used to identify market groupings based on customer behavior , find patterns in stock market data
and assess risk in investment portfolios.

 K-means clustering is a very popular clustering algorithm which applied when we have a dataset with labels
unknown.
 The goal is to find certain groups based on some kind of similarity in the data with the number of groups
represented by K.
 This algorithm is generally used in areas like market segmentation, customer segmentation, etc. But, it can
also be used to segment different objects in the images on the basis of the pixel values.

The algorithm for image segmentation works as follows:

 First, we need to select the value of K in K-means clustering.

 Select a feature vector for every pixel (color values such as RGB value, texture etc.).
 Define a similarity measure b/w feature vectors such as Euclidean distance to measure the similarity b/w
any two points/pixel.
 Apply K-means algorithm to the cluster centres .
 Apply connected component’s algorithm.