ML4EA

Part 2 · Chapter 10

Unsupervised Learning

Introduces methods for discovering structure without labels, including clustering, dimensionality reduction, density estimation, and anomaly detection.

Chapter map

What this chapter develops

  • K-means and hierarchical clustering
  • DBSCAN
  • PCA
  • t-SNE and UMAP
  • Autoencoders
  • Density estimation and anomaly detection

Engineering practice

Application and Diamond Examples

  1. 01

    Hydraulic operating-state clustering

  2. 02

    Power-grid sensor fault detection with DBSCAN

  3. 03

    Vibration fault diagnosis with PCA

  4. 04

    Structural health monitoring with autoencoders

Companion materials

Portal resources

Chapter catalogFind this chapter’s indexed examples and related topics.Application Example catalogOpen 9 validated notebooks with a verified portal account.AE datasetsReview dataset sources and generation details for this chapter's Application Examples.DiscussionAsk chapter-specific questions and compare engineering interpretations with other readers.