Vertex AI Feature Store
In ML/DL model training, features are the attributes to build a model and make future inferences. Google Vertex AI Feature Store is a fully managed cloud service that provides a centralized repository to store, organize, and serve ML features. You can create and manage a Vertex AI Feature Store that contains all the model features and their values. With a central feature store, users in a Google Cloud organization can share and reuse these features for model training or serving tasks in different ML/DL projects to speed up machine learning application development and model deployment.
Vertex AI Feature Store enables users to manage features in ML models. A feature store can serve real-time, online predictions, or batch predictions for the new data. For example, after loading information about movie-watching habits, the Feature Store can serve to predict what movie a new user may watch based on their characteristics.
In traditional ML frameworks, you may...
