When researchers train machine learning or deep learning models, huge amounts of data and multilayer algorithms make the model training slow. AWS EMR services combine with Spark based memory processing architecture that allows users to do this work in a very short time. In addition, H2o is a scalable machine learning framework developed by h2o.ai and supports many APIs (such as R, Python, Scala, and Java, or spark) in statistical languages. Sparkling water is a machine learning framework that is a combination of H2o and Spark, allowing users to use H2o flow UI training and is very suitable for users who are first in contact with machine learning. This experiment mainly leads you to use the R studio in AWS EMR services and Sparkling water to train several machine learning models and to present the results by R flexdachboard package.
http://docs.h2o.ai/h2o/latest-stable/h2o-docs/save-and-load-model.html