AWS Certified Machine Learning – Specialty — Question 229

A machine learning (ML) specialist is developing a deep learning sentiment analysis model that is based on data from movie reviews. After the ML specialist trains the model and reviews the model results on the validation set, the ML specialist discovers that the model is overfitting.

Which solutions will MOST improve the model generalization and reduce overfitting? (Choose three.)

Answer options

• A. Shuffle the dataset with a different seed.
• B. Decrease the learning rate.
• C. Increase the number of layers in the network.
• D. Add L1 regularization and L2 regularization.
• E. Add dropout.
• F. Decrease the number of layers in the network.

Correct answer: D, E, F

Explanation

The correct answers D, E, and F are effective strategies to combat overfitting. Adding L1 and L2 regularization (D) helps penalize complex models, dropout (E) randomly drops units during training to prevent co-adaptation, and decreasing the number of layers (F) simplifies the model. In contrast, options A, B, and C do not directly address the overfitting issue.