AI Glossary

Underfitting

What does “Underfitting” mean?

Underfitting: When a model is too simple or has not trained enough, it misses the real patterns in the data and performs poorly even on its own training set. Common fixes are using a larger model, training longer or giving it better input features.

Underfitting visual guide: When a model is too simple or has not trained enough, it misses the real patterns in the data and performs poorly even on its own training set. Common fixes are using a larger model, training longer or giving it better input features.
Visual guide Open full size (new tab)

What is an example of “Underfitting”?

Fitting a straight line to data that clearly curves will underfit: the line misses the trend no matter how carefully it is placed.

Keep learning

See also

OverfittingTermRegularizationTermLoss functionTermEpochTerm

Same topic

Validation and test setsTermTransfer learningTermSynthetic dataTermScaling lawsTermReward modelTerm