MLA-C01 Question 115
Select 4You are developing a machine learning model for predicting housing prices using Amazon SageMaker. During training, you notice that the validation error is significantly higher than the training error, indicating overfitting. Which of the following actions should you take to address the overfitting issue?
- A
Increase the size of the training dataset by augmenting it or collecting more data
- B
Reduce the complexity of the model architecture by decreasing the number of layers or nodes
- C
Use early stopping during training to prevent the model from overfitting
- D
Increase the learning rate to ensure faster convergence of the model
- E
Add dropout regularization to the model during training
Show answer and explanation
Correct answers: A, B, C, E
Explanation
Overfitting occurs when the model performs well on the training data but poorly on unseen data. To address this, you can increase the size of the training dataset, reduce model complexity, use regularization techniques like dropout, or apply early stopping to prevent overfitting during training. Increasing the learning rate is not a valid solution for addressing overfitting; it primarily affects the speed and stability of training.
- A. Correct.
Increasing the size of the training dataset can help the model generalize better and reduce overfitting by providing more diverse examples.
- B. Correct.
Reducing model complexity (e.g., fewer layers or nodes) can prevent the model from memorizing the training data and help it generalize better.
- C. Correct.
Early stopping is a technique where training halts when the validation error stops improving, which helps mitigate overfitting.
- D. Incorrect.
Increasing the learning rate is not a solution for overfitting. Instead, it may cause the model to converge too quickly or fail to converge properly.
- E. Correct.
Adding dropout regularization introduces noise during training by randomly disabling a fraction of neurons, which reduces overfitting by preventing the model from relying too heavily on specific neurons.