AI & Machine Learning · Foundational

Overfitting

When a model learns training data too closely, including noise, and performs poorly on new data.

Definition

An overfit model has low training error but high validation or test error, which indicates high variance. Causes include too complex a model, too little data or training too long. Remedies include more data, regularization, dropout, early stopping, simpler models, data augmentation and cross-validation.

Don't confuse it with

Related terms