Decoder. plain-English AI glossary

Accuracy

● Core

The fraction of predictions the model got right out of the total—the simplest measure of success.

Think of it like

Like the percentage of test questions a student answers correctly.

Example

If the model classifies 950 emails correctly out of 1000, accuracy is 95%.

How it actually works

Accuracy is intuitive but misleading when classes are imbalanced. If 99% of emails are not spam and your model predicts 'not spam' for everything, it hits 99% accuracy but catches zero spam. Use accuracy for balanced problems or as one metric among many.

For product teams

Report accuracy, but ask what it means. On imbalanced tasks, precision and recall matter more.

For engineers

Always pair accuracy with per-class metrics; plot confusion matrices.

F1 Score — Better for imbalanced data.
Precision — Complementary metric.

Read anything AI without the jargon

Look up any term in plain English, or save terms as you read with the free Chrome extension.

Open Decoder Add to Chrome