Explanation

Conceptual guides to understand the theory and make informed decisions.

---

Core Concepts

• Bregman Divergences — The mathematical foundation
• When to Use What — Decision framework for divergences

Architecture

• How Lloyd’s Algorithm Works — The core iteration
• Assignment Strategies — Cross-join vs broadcast

Performance

• Performance Tuning — Scaling to billions of points
• Acceleration Techniques — Elkan, Hamerly, mini-batch
• Comparison with MLlib — When to use this vs. built-in

Advanced Topics

• Soft vs Hard Clustering — Probabilistic assignments
• Cluster Validity — Evaluation metrics

---

Back to Home