Without labels, how do you convince a stakeholder your clustering is any good?
mediumAnswer
- Combine three kinds of evidence, because no single one is sufficient.
- Internal measures such as silhouette or Davies-Bouldin say whether clusters are compact and separated, which is necessary but not meaningful on its own.
- Stability says whether the structure is real: re-run on bootstrap samples or different seeds and measure how often pairs of points stay together, since a partition that changes every run is describing noise.
- Then external usefulness, which is what actually persuades anyone: profile each cluster on variables you did not cluster on, and show that the groups differ in ways the business recognizes, or that using the cluster as a feature improves a downstream model.
- Present the clusters as hypotheses to be validated, never as ground truth.
Check yourself — multiple choice
- Report inertia only
- Combine internal separation measures, stability across resamples, and external validation on held-out variables or a downstream task
- Clustering cannot be evaluated
- Ask the stakeholder to label the data
Separation, stability and external usefulness together make a defensible case; each alone is weak.
#clustering#evaluation
Practise Unsupervised Learning
214 interview questions in this topic.