AI-900 Question 77
Select 3You are tasked with building a machine learning model using Azure Machine Learning to predict customer churn for a subscription service. You want to ensure the model is trained effectively and performs well. Which of the following are fundamental principles you should consider during the machine learning lifecycle on Azure?
- A
Ensure the data is clean, well-prepared, and relevant to the problem.
- B
Use a single algorithm without experimentation to save time.
- C
Split the data into training and validation datasets to evaluate model performance.
- D
Leverage automated machine learning (AutoML) to test multiple algorithms and hyperparameters.
- E
Ignore the need for feature engineering as Azure handles this automatically.
Show answer and explanation
Correct answers: A, C, D
Explanation
In machine learning on Azure, it is essential to follow key principles such as cleaning and preparing your data, splitting data for training and validation, and leveraging tools like AutoML to optimize model selection. Ignoring these principles can lead to poor model performance, reduced accuracy, and wasted resources.
- A. Correct.
Ensuring data is clean, well-prepared, and relevant is a fundamental principle of machine learning. Without quality data, the model will not perform effectively.
- B. Incorrect.
Using a single algorithm without experimentation is not recommended. Experimentation with different algorithms and hyperparameters is necessary to identify the best-performing model.
- C. Correct.
Splitting the data into training and validation datasets is critical for evaluating model performance and avoiding overfitting.
- D. Correct.
Azure's AutoML can automate the process of testing multiple algorithms and hyperparameters, helping to identify the best model for the problem efficiently.
- E. Incorrect.
Feature engineering is still important in machine learning. While Azure provides tools to assist, ignoring this step entirely can lead to suboptimal model performance.