Adjusted Rand Index vs NMI — external cluster metrics.
mediumAnswer
- Both compare a clustering to ground-truth labels (when available).
- ARI = (RI - E[RI]) / (max RI - E[RI]) → 0 for random, 1 for perfect match, can be negative.
- Corrects for chance.
- NMI = mutual information normalized to [0,1] — measures how much info the clustering carries about labels.
- AMI = adjusted NMI (chance-corrected).
- Rule: report both — they disagree in edge cases.
Check yourself — multiple choice
- Same thing
- ARI: chance-corrected Rand, 0=random / 1=perfect; NMI: normalized mutual info in [0,1] — report both external metrics
- Only ARI
- Random
ARI + NMI (or AMI): standard external clustering evaluation.
#clustering#evaluation
Practise Unsupervised Learning
214 interview questions in this topic.