Unsupervised Learning Notes - Class 12 Data Science (844)

Get Unsupervised Learning Notes specially desinged for Data Science students of Class 12 as per CBSE latest Curriculum.

Contents hide

Introduction to Unsupervised Learning

In Supervised Learning, algorithms are trained using well-labeled data, where the correct labels are already available. The trained model is then used to make predictions for classification or regression problems.
Unsupervised Learning is a machine learning approach in which algorithms are not trained using labeled or classified data.
In Unsupervised Learning, algorithms work on the data without human intervention to analyze it.
The main objective of Unsupervised Learning is to discover hidden patterns or groupings in the data by identifying similarities and differences among data points.
Unsupervised Learning is widely used in real-life applications such as exploratory data analysis, customer segmentation, and image recognition.

Recommendation Engines: Unsupervised learning is used by many e-commerce websites to identify patterns in customers’ past purchase behavior and recommend relevant products during the checkout process.
Medical Imaging: Unsupervised learning supports medical imaging by performing tasks such as image detection, classification, and segmentation, helping doctors diagnose patients quickly and accurately.
Anomaly Detection: Unsupervised learning analyzes large amounts of data to identify unusual data points (anomalies), which can help detect fraud in financial systems and security breaches.
Customer Personas: Unsupervised learning helps organizations identify common purchasing habits and buying times of customers, enabling them to create better customer profiles and target sales and advertisements more effectively.
News Sections: Online news websites use unsupervised learning to automatically categorize news articles into different sections such as sports, entertainment, and international news.

Clustering is the process of dividing raw data into different groups (clusters) so that data points within the same group are similar to each other, while data points in different groups are different.
Clustering is used when there is no class to be predicted. Instead, the objective is to divide the data into natural groups based on their similarities.
In clustering, the input is the original raw data, and the output is a dataset divided into distinct clusters.
A real-life example of clustering is the arrangement of products in a supermarket, where similar items, such as vegetables and fruits, are placed in separate sections.
The main clustering methods are:
- Partitioning Clustering
- Density-Based Clustering
- Distribution Model-Based Clustering
- Hierarchical Clustering

K-Means Clustering is an unsupervised machine learning technique used to identify clusters (groups) of similar data points in a dataset.
Among the different clustering methods, K-Means is one of the simplest and easiest algorithms to understand and implement.
The algorithm starts by randomly selecting centroids, where the number of centroids is equal to the number of clusters (K) chosen by the user.
K-Means follows the Expectation-Maximization approach to group similar data points into clusters.
The steps of the K-Means algorithm are:
1. Specify the number of clusters (K).
2. Initialize the centroids by randomly selecting K data points.
3. Repeat the process until the centroids no longer change and the assignment of data points to clusters remains the same.
4. Compute the sum of the squared distance between the data points and all centroids.
5. Assign each data point to the cluster whose centroid is closest to it.
K-Means Clustering is widely used in real-world applications such as market segmentation, document clustering, and image segmentation.