When launching an Amazon SageMaker training job with a built-in algorithm, which parameters are required to be provided? (Choose three.)

When launching an Amazon SageMaker training job, three fundamental parameters are always required. First, a training channel is essential to specify the Amazon S3 location where your training data resides, as the algorithm needs data to learn from. Second, an IAM role is mandatory. SageMaker operates on your behalf to access S3, launch instances, and perform other tasks, and it needs explicit permissions granted via an IAM role. Third, an output path specifying an S3 location is required to store the trained model artifacts once the training job completes. A validation channel is optional; while highly recommended for model evaluation, training can proceed without it. Hyperparameters are often required but not always; some built-in algorithms have default hyperparameters, making them optional for initial runs. The EC2 instance type is also required, but the question asks for parameters, and the instance type is typically specified as part of the resource configuration, not a core parameter for the job itself in the same way as data, role, and output.