In Machine Learning, what is “overfitting”?
a) When a model learns the training data too well, including noise
b) When a model generalizes perfectly to new data
c) When a model underperforms on both training and test data
d) When a model is too simple for the data
Answer:
a) When a model learns the training data too well, including noise
Explanation:
Overfitting occurs when a machine learning model becomes too complex and fits the training data too closely. This includes learning not only the underlying patterns but also the noise and fluctuations in the training data, resulting in poor performance on new, unseen data.
To prevent overfitting, techniques such as cross-validation, regularization, and pruning are used. These methods help the model generalize better by reducing its complexity and preventing it from memorizing the training data.
A well-balanced machine learning model should perform well on both training data and unseen data, indicating that it has captured the essential patterns while ignoring irrelevant details.