Microsoft Azure AI Fundamentals (AI-900) Practice Exam

What does Automated ML enable users to specify for predictions?

1. A dataset for training only

2. A dataset and label to predict

3. Only the labels to predict

4. Predefined features and parameters

The correct answer is: A dataset and label to predict

Automated ML in Azure enables users to specify both a dataset and a label to predict, which is fundamental for supervised machine learning tasks. When you select a dataset, you provide the raw input data that will be used for training the model. The label, on the other hand, represents the target variable or outcome that you want the model to predict based on the input features in your dataset. This combination allows Automated ML to understand what to learn from the data and what predictions to make, streamlining the process of model creation and enabling non-experts to utilize machine learning effectively. In contrast, specifying only a dataset for training would limit the model's ability to understand what it is being trained to predict. Providing only the labels to predict without the corresponding dataset would not furnish the model with the necessary context or input data for making accurate predictions. Finally, focusing solely on predefined features and parameters overlooks the flexibility of including the entire dataset and target label for more robust model training. Thus, the option of specifying both a dataset and a label to predict is crucial for the effectiveness of Automated ML processes.