Databricks Machine Learning Associate Question 231
Select 3In a Databricks machine learning workflow, you are training a model to predict customer churn. You decide to use cross-validation instead of a simple train-validation split. Which of the following are advantages and disadvantages of using cross-validation in this scenario?
- A
Cross-validation provides a more reliable estimate of model performance by reducing the impact of data splits.
- B
Cross-validation is computationally more expensive compared to a train-validation split, especially with large datasets.
- C
Cross-validation always prevents overfitting, regardless of the model or dataset.
- D
Cross-validation ensures that the entire dataset is used for both training and validation at different points.
- E
Cross-validation reduces training time compared to a simple train-validation split.
Show answer and explanation
Correct answers: A, B, D
Explanation
Cross-validation is a technique used to evaluate model performance by splitting the dataset into multiple folds and training the model on different subsets of the data. Its key advantages include providing a more reliable estimate of model performance by reducing the impact of individual data splits and ensuring the entire dataset is utilized for training and validation. However, it comes with the trade-off of higher computational cost, as the model needs to be trained multiple times. While it aids in assessing generalization, it does not inherently prevent overfitting or reduce training time.
- A. Correct.
Correct: Cross-validation provides a more robust evaluation of model performance by repeatedly splitting the dataset and averaging the results, mitigating bias due to a single train-validation split.
- B. Correct.
Correct: Cross-validation requires training the model multiple times (once per fold), which increases computational cost, especially for large datasets or complex models.
- C. Incorrect.
Incorrect: While cross-validation helps in assessing model generalization, it does not inherently prevent overfitting. Overfitting is more related to factors like model complexity and hyperparameter selection.
- D. Correct.
Correct: Cross-validation ensures that every data point is used in both training and validation at different iterations, which can improve the reliability of evaluation metrics.
- E. Incorrect.
Incorrect: Cross-validation increases computational time since it requires multiple training iterations, making it slower than a simple train-validation split.