Overfitting

Sat, 18 Jul 2026 09:41:54 +0000

Definition

Overfitting occurs when a model learns the training data too well, including its random noise and outliers, resulting in excellent performance on training data but poor performance on new, unseen test data. This happens because the model becomes overly complex relative to the amount of training data available. Techniques like regularization, dropout, early stopping, and cross-validation are commonly employed to mitigate overfitting and improve the model’s ability to generalize.

Model Evaluation on English AI Terms Dictionary

Overfitting

Definition