📏 One-vs-Rest (OvR)

One-vs-Rest (OvR)#

A multiclass strategy fitting one binary classifier per class.

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 Despite careful review, it may still contain mistakes, omissions, or out-of-date information.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#

One-vs-Rest (also one-vs-all) reduces a K-class problem to K binary ones — in each, a single class is the positive and all the others are lumped together as the negative. It’s the simplest way to let binary tools handle many classes.

How it’s used#

For a K-class model you get K ROC curves and AUCs, one per class, each answering how well does the model separate this class from everything else? scikit-learn exposes it as multi_class='ovr'; it also matches the multilabel setting, where classes aren’t exclusive.

The catch#

Each binary split is imbalanced — the positive class is only about 1/K of the data, and the “rest” group’s makeup shifts with the class distribution, so OvR scores are sensitive to class imbalance. The alternative, One-vs-One, compares class pairs and is less imbalance-prone but trains \(O(K^2)\) classifiers.

Theme: Classification & Averaging Metrics · All terminology

Hint

Mind map — connected ideas

Multiclass Classification · Macro AUC · Micro AUC · Binary Classification · ROC-AUC (Receiver Operating Characteristic – Area Under Curve, = AUROC) · Multiclass AUROC

Hint

More in Classification & Averaging Metrics

Accuracy · AUC (Area Under the Curve) · Average Precision (AP) · Binary Classification · Classification Probability · Discriminatory Power · F1-score · Gini Coefficient · Harmonic Mean · Log Loss (also called Logarithmic Loss or Cross-Entropy Loss) · Macro AUC · Macro AUROC (Macro-Averaged AUROC) · Macro Averaging · Macro F1