Overfit

Quick meaning: Overfit is a verb meaning (statistics) To use a statistical model that has too many parameters relative to the size of the sample, leading to a good fit with the sample data but a poor fit with new data.

Meaning

(statistics) To use a statistical model that has too many parameters relative to the size of the sample, leading to a good fit with the sample data but a poor fit with new data.

Example Sentences

If your strategy is to always fit and use the global model, you will probably overfit the model (i.e., include unnecessary variables).

In all modeling techniques, and neural networks in particular, care must be taken not to overtrain or overfit the model. If possible, models should be interpreted in terms of their mechanistic meaning.

Word Forms

• overfits
• overfitting
• overfitted

Antonyms for overfit

Common opposites include underfit.

Related Words

Related dictionary entries include overparameterize, overparameterise, overparametrize, overspecify, overoptimize, overextrapolate.

Related Dictionary Tools