Adding a Dataset to Driverless AI from Amazon S3
- Predictive AI
- H2O Driverless AI
This lesson walks through the Datasets tab in Driverless AI, where every dataset added to the platform is listed with its name, path, size, row and column counts, status, and creation date. It covers the different sources a dataset can be added from, including the local file system, Amazon S3, H2O Drive, Snowflake, or the Feature Store, and then demonstrates importing a Credit_Score.csv file from H2O's public Amazon S3 bucket. It also notes that datasets persist for a regular account but are erased after two hours on Aquarium, and shows how to search or delete a dataset from the list.
Key takeaways
- How to import a dataset into Driverless AI from Amazon S3 by pointing to the H2O public bucket and selecting a file
- What information the Datasets tab shows for each file: its path, size, row and column counts, status, and creation date
- Why datasets persist for a regular Driverless AI account but get erased after two hours when using Aquarium
- The other sources a dataset can be added from, including the local file system, H2O Drive, Snowflake, and the Feature Store
- How to search for a specific dataset by name or date, or delete one by selecting its checkbox
Start owning your AI