Feature store using Feast
In order to ensure effective development throughout the machine learning lifecycle, it is important to maintain detailed and comprehensive records of resources. Feature stores and model registries are examples of helpful resource records in the pre-modelling and modelling phases. In this exercise, you will implement a feature store using Feast. The predefined patient, Entity, as well as the cp, thalach, ca, and thal features have been loaded for you. ValueType, FeatureStore, and FileSource are all imported from feast. heart_disease_df is also imported.
Cet exercice fait partie du cours
End-to-End Machine Learning
Exercice interactif pratique
Essayez cet exercice en complétant cet exemple de code.
heart_disease_df.to_parquet("heart_disease.parquet")
# Point File Source to the saved file
data_source = ____(
path=____,
event_timestamp_column="timestamp",
created_timestamp_column="created",
)