Overfitting

Category: science

When a model learns the "noise" in training data so well that it fails to generalize to new, unseen data.

Overfitting is the "memorization trap." The model performs perfectly on the test data because it has essentially memorized every specific outlier, but it falls apart when you try to use it on real-world, live production traffic.

Common Examples

• The diagnostic report showed that our model was heavily overfitting, leading to high variance when applied to the new market dataset.
• We introduced "dropout" layers to our neural network to combat overfitting and ensure our predictions remain robust for unseen cases.