Source Of Summary:

Lec.3.D. M. FCDS spring 2024.pdf

Data Mining and Analytics – Lecture 3

1. Introduction to Hierarchical Clustering

Hierarchical clustering groups data objects into a tree-like structure (a dendrogram) where every element finds its perfect spot like a family tree for your data! 👨‍👩‍👧‍👦🎉

1.1 What Is It? 🤔

• Goal:

  Create a hierarchy (dendrogram) from an input set $S$ so that:

  • Root: Represents the whole set $S$ 🌐
  • Leaves: Represent individual data points 🍃
  • Internal Nodes: Formed by merging their child clusters 🤝

• Visualization:

  Each level partitions your data into nested clusters like peeling an onion layer by layer! 🧅😆

1.2 Types of Hierarchical Clustering 🛠️✨

1.2.1 Agglomerative (Bottom-Up) 👤➡️

• Start: Each data point begins in its own cluster (a singleton) 👤
• Process: Merge the closest clusters step by step until the whole set $S$ becomes one big cluster! 😄💪
• Fun Fact: Think of it like rolling a snowball that gets bigger and bigger! ☃️❄️

1.2.2 Divisive (Top-Down) 🍰✂️

• Start: Begin with the entire set $S$ 🏁
• Process: Recursively split the set into smaller clusters until each cluster contains a single element.
• Note: It’s like slicing a cake into individual pieces one careful cut at a time! 🍰✂️

2. Agglomerative Clustering Algorithm 🤖➡️