Databricks Certified Machine Learning - Associate
Get started today
Ultimate access to all questions.
Ultimate access to all questions.
A data scientist is optimizing hyperparameters using an iterative optimization algorithm, with each unique set of hyperparameters trained on a separate compute node. Despite conducting eight evaluations across eight nodes, there's no consistent improvement in model accuracy. What single change could most effectively enhance the model's accuracy during the tuning process?
Explanation:
B. Change the iterative optimization algorithm to one better suited for the task.
Explanation:
The lack of consistent improvement in model accuracy suggests the current optimization algorithm may not be effectively exploring the hyperparameter space for this specific problem. Different algorithms have unique strengths and may perform better depending on the dataset or problem type. Therefore, switching to a more suitable optimization algorithm could lead to better hyperparameter selection and improved model accuracy.
Other options focus on adjusting computational resources (A, C, D, E), which do not directly address the effectiveness of the hyperparameter search strategy. The key to enhancing accuracy lies in the algorithm's ability to efficiently navigate the hyperparameter space, not merely in scaling the number of evaluations or compute nodes.