MLS-C01 Question 224
Select 4You are building a machine learning model to classify customer feedback as positive, neutral, or negative. During model evaluation, you observe high training accuracy but significantly lower validation accuracy. Which of the following strategies can you apply to address this issue?
- A
Reduce the size of your model to decrease its capacity.
- B
Increase the size of the training dataset by collecting more labeled data.
- C
Use regularization techniques such as L1 or L2 regularization.
- D
Increase the number of epochs in your training process.
- E
Apply data augmentation techniques to diversify the training data.
Show answer and explanation
Correct answers: A, B, C, E
Explanation
The issue described in the scenario indicates that the model is overfitting to the training data, which is evident from the high training accuracy and low validation accuracy. Strategies like reducing model capacity, increasing training data, applying regularization, and using data augmentation are effective in reducing overfitting. However, increasing the number of epochs may worsen overfitting and is not a suitable solution in this case.
- A. Correct.
Reducing the size of the model can help prevent overfitting by lowering its capacity to memorize the training data, which can improve validation performance.
- B. Correct.
Increasing the size of the training dataset provides more examples for the model to learn from, reducing overfitting and improving generalization.
- C. Correct.
Using regularization techniques such as L1 or L2 regularization helps penalize large weights in the model, preventing overfitting and improving generalization.
- D. Incorrect.
Increasing the number of epochs can exacerbate overfitting if the model starts memorizing the training data, which would not help in this scenario.
- E. Correct.
Applying data augmentation techniques diversifies the training data, making the model more robust and less likely to overfit.