No-Code Machine Learning Using Amazon AWS SageMaker Canvas - Versioning

The video tutorial explains how to create and manage different versions of a project using SageMaker Canvas. It uses a banknote authentication project as an example, demonstrating how to modify data columns and observe changes in model accuracy. The tutorial emphasizes that while you can change data columns, the target column must remain the same. It highlights the importance of versioning in comparing model performance and improving project outcomes.

Why is versioning important in a machine learning project?

• To manage changes in the dataset

• To create a new project

• To change the target column

• To increase the number of columns

What remains unchanged when creating a new version of a project?

• The accuracy

• The number of columns

• The target column

• The dataset

In the V3 version of the project, what was the observed accuracy after removing certain columns?

• 90.0%

• 93.9%

• 95.0%

• 98.6%

What is the main benefit of comparing different versions of a model?

• To change the target column

• To find the best performing model

• To increase the number of versions

• To reduce the dataset size

What is a key restriction when creating new versions in SageMaker Canvas?

• You cannot change the target column

• You cannot add new columns

• You cannot remove columns

• You cannot change the dataset