📏  Macro AUC

Macro AUC#

AUC averaged equally across per-class scores.

Important

✨ AI-generated content. This page was written with the assistance of an AI language model and is provided as a learning aid. Despite careful review, it may still contain mistakes, omissions, or out-of-date information. Whether you are new to the topic, a team lead, or a senior practitioner, treat it as a starting point rather than an authoritative reference: read it critically and independently verify anything you act on (code, commands, figures, and factual claims) against official documentation and primary sources before relying on it.

What it is#

Macro AUC averages the per-class AUCs (from One-vs-Rest) with equal weight — every class counts the same, no matter how rare or common. It answers how well does the model do on the average class?

Macro vs micro#

The contrast is micro AUC, which pools every class’s true/false-positive contributions into one global curve, effectively weighting by prevalence so frequent classes dominate. Macro treats a class with 10 samples exactly like one with 10,000.

When to use which#

Macro is the choice when the rare classes matter as much as the common ones (you want minority performance to show), while micro (or a weighted macro) better reflects overall accuracy on imbalanced data. Reporting both reveals whether a good score is carried by the majority classes.


Theme: Classification & Averaging Metrics  ·  All terminology



See also

Source article Adapted (context, re-expressed) in our own words from: Macro AUC (insightful-data-lab.com).

Tags: purpose: reference topic: terminology level: intermediate