When a data scientist integrates a random forest regressor pipeline as the final stage in a Spark ML Pipeline and initiates cross-validation, what is a potential downside of constructing the pipeline within the cross-validation process?

Real Exam

The process might not be able to parallelize tuning because of the pipeline's distributed nature.

6.3%

There's a risk of leaking data preparation details from validation sets to training sets for each model.

18.8%

The runtime could significantly increase as every pipeline stage requires refitting or retransforming for each model iteration.

68.8%

It may fail to evaluate all unique hyperparameter value combinations in the parameter grid.

6.3%

Data leakage from the training set to the test set could occur during evaluation.

Databricks Certified Machine Learning - Associate