Module 2: The Discovery Branch
Finding Groups You Didn't Know Existed
Learning Objectives
- Sort innovation projects into natural groups without labels
- Apply K-Means clustering to group by similarity
- Use PCA and UMAP for dimensionality reduction
- Interpret cluster profiles as innovator archetypes
- Validate whether discovered groups are real
Topics
| Topic | Type |
|---|---|
| Decision Tree Navigator | Opening |
| Are There Distinct Types? | Hook |
| K-Means: Grouping by Similarity | Lecture |
| K-Means: Grouping by Similarity | Lecture |
| Seeing the Big Picture: Dimensionality Reduction | Lecture |
| Clustering Demo | Live Demo |
| Cluster Profiles | Lecture |
| Are the Groups Real? Validation | Lecture |
| Reporting Clustering | Lecture |
| Your Turn: Where Does This Fit? | Socratic |
| Map Check: Where Are We? | Discussion |