What is overfitting?

What is overfitting?

• A. When a model memorizes training data and performs poorly on new data
• B. When a model underestimates complexity
• C. When a model achieves perfect accuracy on test
• D. When a model has high bias

Answer: (C)

Overfitting happens when a model learns the training data too well, including its noise, so it fits the training set perfectly but generalizes poorly to new, unseen data. The key issue is generalization: a model that overfits has high accuracy on the data it was trained on but low accuracy on new data because it hasn’t captured the underlying patterns, just the quirks of the training set. Getting perfect accuracy on a test set would not indicate overfitting; it would suggest the model generalizes extremely well or that some leakage or an unusually easy test set is present. Conversely, underfitting (high bias) occurs when the model is too simple to capture the patterns, leading to poor performance on both training and test data. Overfitting is about memorizing the training data and failing to generalize, not about achieving perfect test accuracy.

Overfitting happens when a model learns the training data too well, including its noise, so it fits the training set perfectly but generalizes poorly to new, unseen data. The key issue is generalization: a model that overfits has high accuracy on the data it was trained on but low accuracy on new data because it hasn’t captured the underlying patterns, just the quirks of the training set. Getting perfect accuracy on a test set would not indicate overfitting; it would suggest the model generalizes extremely well or that some leakage or an unusually easy test set is present. Conversely, underfitting (high bias) occurs when the model is too simple to capture the patterns, leading to poor performance on both training and test data. Overfitting is about memorizing the training data and failing to generalize, not about achieving perfect test accuracy.