Databricks Certified Machine Learning Associate — Question 22

A data scientist wants to parallelize the training of trees in a gradient boosted tree to speed up the training process. A colleague suggests that parallelizing a boosted tree algorithm can be difficult.
Which of the following describes why?

Answer options

• A. Gradient boosting is not a linear algebra-based algorithm which is required for parallelization.
• B. Gradient boosting requires access to all data at once which cannot happen during parallelization.
• C. Gradient boosting calculates gradients in evaluation metrics using all cores which prevents parallelization.
• D. Gradient boosting is an iterative algorithm that requires information from the previous iteration to perform the next step.
• E. Gradient boosting uses decision trees in each iteration which cannot be parallelized.

Correct answer: D

Explanation

The correct answer is D because gradient boosting relies on the results from the previous iteration to inform the next one, making it inherently sequential. The other options either misrepresent how gradient boosting operates or incorrectly describe the requirements for parallelization.