majiayu000 Clustering Algorithm Runner
Use this skill when the main question is how to group unlabeled data points.
Overview
This skill covers algorithm choice, preprocessing implications, cluster validation, and interpretation for unsupervised segmentation problems.
When to Use This Skill
- Customer segmentation, cohort discovery, or grouping unlabeled records
- Choosing between centroid, density, or hierarchical clustering
- Reviewing silhouette score, Davies-Bouldin, or cluster stability
Not For / Boundaries
- Supervised prediction with labels: use
training-machine-learning-models - Pure anomaly review without clustering as the central method: use
anomaly-detector - Final narrative report packaging: use
scientific-reporting
Typical Outputs
- Algorithm recommendation with parameter guidance
- Cluster-assignment workflow
- Validation and interpretation notes for cluster quality
Related Skills
creating-data-visualizationsfor exploratory plots of cluster structureanomaly-detectorwhen outliers become the next question