
Ultimate access to all questions.
Your company operates a global e-commerce platform that leverages TensorFlow-based deep learning models for processing Analytics-360 data to classify various business outcomes. These models, primarily utilizing customers and orders data, have been in production for several months. Recently, your Manager identified an issue where different teams across various projects inconsistently handle the same features derived from the same dataset, leading to unexpected model drift over time. This inconsistency not only affects model performance but also complicates maintenance and scalability efforts. To ensure consistency, reduce redundancy, and improve model reliability, what strategy would you recommend to your Manager? (Select 2 options)
A
Implement a centralized feature repository by searching for and reusing features stored in Cloud Storage.
B
Utilize Vertex Feature Store to insert or update features, ensuring a centralized and consistent feature management system.
C
Encourage each team to classify their features independently and share them manually with other teams.
D
Search the Vertex Feature Store for existing features to reuse, avoiding duplication and ensuring consistency.
E
Store each model's features separately in Cloud Storage, creating isolated feature sets for each project.