Mitigating underfitting and overfitting: Practice Questions — Data Science Pipelines and Workflow Automation (NVIDIA-Certified Associate: Accelerated Data Science)

Practice Questions: Mitigating Underfitting and Overfitting

These multiple-choice questions are designed to help you prepare for the NVIDIA-Certified Associate: Accelerated Data Science exam, focusing on the critical topic of mitigating underfitting and overfitting within data science pipelines.

1. Which of the following techniques is most effective at reducing overfitting in a machine learning model?

   • A. Increasing the number of features without selection
   • B. Applying regularization methods such as L1 or L2 penalties
   • C. Decreasing the size of the training dataset
   • D. Using a simpler model with fewer parameters

   Correct answer: B

   Explanation: Regularization methods like L1 and L2 add penalties to large coefficients, discouraging complex models that overfit the training data. While simpler models (D) can reduce overfitting, regularization is a more targeted approach. Increasing features (A) or reducing data (C) can worsen overfitting.

2. What is a common symptom of underfitting in a model's performance?

   • A. High training accuracy but low validation accuracy
   • B. Low training accuracy and low validation accuracy
   • C. High validation accuracy but low training accuracy
   • D. High training and validation accuracy

   Correct answer: B

   Explanation: Underfitting occurs when the model is too simple to capture underlying patterns, resulting in poor performance on both training and validation sets.

3. Which approach helps to mitigate both underfitting and overfitting by improving feature representation?

   • A. Feature engineering and transformation
   • B. Reducing the number of training epochs drastically
   • C. Using only raw input features without preprocessing
   • D. Ignoring feature correlation

   Correct answer: A

   Explanation: Feature engineering and transformation enhance the model's ability to learn relevant patterns, helping to avoid underfitting and overfitting by providing better input data.

4. Which of the following is a valid method to detect overfitting during model training?

   • A. Training accuracy decreases while validation accuracy increases
   • B. Training accuracy and validation accuracy both increase steadily
   • C. Training accuracy is high but validation accuracy plateaus or decreases
   • D. Both training and validation accuracy remain low

   Correct answer: C

   Explanation: Overfitting is indicated when the model performs very well on training data but poorly or stagnantly on validation data, showing it has memorized training examples rather than generalized.

5. How can dataset augmentation help in mitigating overfitting?

   • A. By reducing the number of samples in the dataset
   • B. By artificially increasing dataset diversity to improve generalization
   • C. By removing noisy data points
   • D. By simplifying the model architecture

   Correct answer: B

   Explanation: Dataset augmentation creates varied examples from existing data, helping models generalize better and reducing overfitting risk.

6. Which workflow automation tool is commonly used with RAPIDS to build reproducible data science pipelines that help manage overfitting risks?

   • A. Apache Airflow
   • B. Dask
   • C. TensorBoard
   • D. Jupyter Notebook

   Correct answer: B

   Explanation: Dask integrates with RAPIDS to enable scalable, reproducible pipelines that facilitate iterative experimentation and monitoring, which are key to controlling overfitting.

7. What is the effect of increasing model complexity without adequate data or regularization?

   • A. It reduces both underfitting and overfitting
   • B. It increases underfitting
   • C. It increases overfitting
   • D. It has no effect on model performance

   Correct answer: C

   Explanation: Increasing model complexity without sufficient data or regularization usually leads to overfitting, where the model captures noise instead of general patterns.

More in this topic

Mitigating underfitting and overfitting — Data Science Pipelines and Workflow Automation (NVIDIA-Certified Associate: Accelerated Data Science)Building reproducible pipelines with RAPIDS and Dask — Data Science Pipelines and Workflow Automation (NVIDIA-Certified Associate: Accelerated Data Science)Data Science Pipelines and Workflow Automation — NVIDIA-Certified Associate: Accelerated Data ScienceMitigating underfitting and overfitting: Common Mistakes — Data Science Pipelines and Workflow Automation (NVIDIA-Certified Associate: Accelerated Data Science)Feature engineering, selection, and transformation: Practice Questions — Data Science Pipelines and Workflow Automation (NVIDIA-Certified Associate: Accelerated Data Science)Mitigating underfitting and overfitting: Quick Reference — Data Science Pipelines and Workflow Automation (NVIDIA-Certified Associate: Accelerated Data Science)Dataset augmentation and integration — Data Science Pipelines and Workflow Automation (NVIDIA-Certified Associate: Accelerated Data Science)Feature engineering, selection, and transformation: Common Mistakes — Data Science Pipelines and Workflow Automation (NVIDIA-Certified Associate: Accelerated Data Science)End-to-end data science pipeline design — Data Science Pipelines and Workflow Automation (NVIDIA-Certified Associate: Accelerated Data Science)Mitigating underfitting and overfitting: Worked Example — Data Science Pipelines and Workflow Automation (NVIDIA-Certified Associate: Accelerated Data Science)Feature engineering, selection, and transformation: Worked Example — Data Science Pipelines and Workflow Automation (NVIDIA-Certified Associate: Accelerated Data Science)Feature engineering, selection, and transformation — Data Science Pipelines and Workflow Automation (NVIDIA-Certified Associate: Accelerated Data Science)Feature engineering, selection, and transformation: Quick Reference — Data Science Pipelines and Workflow Automation (NVIDIA-Certified Associate: Accelerated Data Science)