Question 15
Domain 2: Explore data and run experimentsYou want to enable early termination in an AutoML classification job to avoid wasting resources on poorly performing trials. Which setting should you configure?
Correct answer: B
Explanation
Configure "enable_early_termination" because it tells AutoML to stop trials that are performing poorly before they consume more compute. Early termination is used to "avoid wasting resources" by ending unpromising runs sooner.
Why each option is right or wrong
A. enable_onnx_compatible_models
B. enable_early_termination
In Azure Machine Learning AutoML classification, the `enable_early_termination` setting on the AutoML configuration is the control that allows the service to stop underperforming child runs before they finish their allotted compute. This is the documented mechanism for pruning poor trials during the experiment, so it directly matches the goal of conserving resources by ending unpromising runs early.
C. trial_timeout_minutes
D. max_trials