AI-900 Question 78
Select 4A retail company wants to create a predictive model to forecast product demand for the next quarter. They have historical sales data stored in an Azure SQL Database. The team decides to use Azure Machine Learning to build the model. Which steps should they include in their machine learning workflow?
- A
Identify and prepare the data for training
- B
Select an appropriate machine learning algorithm
- C
Deploy the trained model to an Azure IoT Edge device
- D
Split the data into training and testing datasets
- E
Generate predictions using the trained model
Show answer and explanation
Correct answers: A, B, D, E
Explanation
The machine learning workflow on Azure involves key steps such as data preparation, selecting an algorithm, training the model, evaluating it, and using it to generate predictions. Deploying to IoT Edge is outside the scope of this scenario as it focuses on demand forecasting, not edge device implementation.
- A. Correct.
Correct. Preparing data is a fundamental step in the machine learning workflow to ensure the data is clean, consistent, and in a usable format for training.
- B. Correct.
Correct. Selecting an appropriate machine learning algorithm is critical to building a model that aligns with the problem being solved (e.g., regression for forecasting).
- C. Incorrect.
Incorrect. Deploying to an Azure IoT Edge device is not relevant to this scenario, as the focus is on creating a predictive model, not IoT implementation.
- D. Correct.
Correct. Splitting the data into training and testing datasets is essential to train the model and evaluate its performance.
- E. Correct.
Correct. Generating predictions using the trained model is the final step in the workflow to apply the model to real-world data.