Build Data Pipelines for Data Science Workloads SNOWPRO-ADVANCED-DATA-SCIENTIST Practice Questions (Page 6)

Questions 26–26

1. 26application · medium

   A data science team wants to train a machine learning model using data stored in Snowflake. They plan to use a Python library that is not available in Snowflake's Python environment. They want to minimize data movement and keep the training data in Snowflake as much as possible. Which approach should they take?

   • AUse Snowpark ML to train the model using the library that is available in Snowflake, or use a Snowpark DataFrame to push down transformations and then train externally.
   • BUse Snowpark to load the data into a pandas DataFrame and train the model using a Python library in a Snowflake stored procedure.
   • CUse Snowflake's external function to call an external ML service that trains the model on the data.
   • DExport the data to a CSV file and load it into a local Python environment for training.
   Reveal Solution Discussion Select an answer first