Summary

In this chapter, we learned how to use H2O AutoML with Apache Spark using an H2O system called H2O Sparkling Water. We started by understanding what Apache Spark is. We investigated the various components that make up the Spark software. Then, we dived deeper into its architecture and understood how it uses a cluster of computers to perform data analysis. We investigated the Spark cluster manager, the Spark driver, Executor, and also the Spark Context. Then, we dived deeper into RDDs and understood how Spark uses them to perform lazy evaluations on transformation operations on the dataset. We also understood that Spark is smart enough to manage its resources efficiently and remove any unused RDDs during operations.

Building on top of this knowledge of Spark, we started exploring what H2O Sparkling Water is and how it uses Spark and H2O together in a seamlessly integrated system. We then dove deeper into its architecture and understood its two types of backends that can be...