The correct answer is B. timeout_minutes. This parameter is specifically designed to set a time limit for AutoML runs, ensuring they don't exceed a specified duration. It's crucial for managing resource usage and scheduling, especially in shared environments or when dealing with time constraints.

• max_trials: Controls the maximum number of model training attempts, but doesn't directly limit time.
• exclude_cols: Specifies columns to exclude from training, but doesn't affect duration.
• pos_label: Identifies the positive class label for binary classification, but has no impact on run time.

Example Usage:

from databricks import automl
automl_classifier = automl.classify(
    target_col="label",
    data=data,
    timeout_minutes=60  # Set a 60-minute timeout
)

from databricks import automl
automl_classifier = automl.classify(
    target_col="label",
    data=data,
    timeout_minutes=60  # Set a 60-minute timeout
)

Key Points:

• Managing Run Duration: timeout_minutes is essential for controlling AutoML run time and preventing excessive resource consumption.
• Resource Management: It's a valuable tool for optimizing resource usage and ensuring efficient model building processes.
• Scheduling Considerations: It's also helpful for accommodating time constraints and fitting AutoML runs into specific schedules.