Question 19
Domain 3: Train and evaluate modelsYou are senior Azure Machine Learning Associate of your company. You are building a recurrent neural network to perform a binary classification. The training loss, validation loss, training accuracy, and validation accuracy of each training epoch has been provided. You need to identify whether the classification model is overfitted. Which of the following is correct?
Correct answer: D
Explanation
Overfitting is identified when a model learns the training data too well but performs worse on unseen data. A common sign is that "the training loss decreases while the validation loss increases" because the model is fitting noise or specifics of the training set rather than general patterns.
Why each option is right or wrong
A. The training loss increases while the validation loss decreases when training the model
B. The training loss stays constant and the validation loss decreases when training the model
C. The training loss stays constant and the validation loss stays on a constant value and close to the training loss value when training the model
D. The training loss decreases while the validation loss increases when training the model
In a binary classification training log, overfitting is indicated when the model keeps improving on the training set but degrades on the holdout set: the training loss falls while the validation loss rises across epochs. That pattern shows the optimizer is still reducing empirical risk on the training data, but generalization error is increasing on unseen data, which is the classic diagnostic for overfitting in epoch-by-epoch monitoring.