Overfitting

Pronunciation: /ˈoʊvɝˈfɪtˈtɪŋ/

Overfitting (noun)

  1. A problem where a model learns too much from training data, failing to perform well on new data.

Examples

  • Overfitting occurs when a model learns the training data too well.
  • Overfitting happens when a model learns too much detail.

Common collocations: prevent overfitting, detect overfitting, reduce overfitting

Synonyms of Overfitting

Formal

Neutral

Antonyms of Overfitting

  • — Describes a model that is too simple to capture the underlying trend.