This image illustrates the client/server architecture of Oracle Machine Learning for R. The components of OML4R are: a client R engine with the OML4R client and other R packages and an Oracle Database instance with OML4R server and possibly other R packages.
The client sends R functions that are translated to SQL commands to the database, which contains data in user tables. The results are returned to the client. The OML4R server can spawn R sessions on the database server for embedded R execution functions, which return results to the OML4R server. The OML4R server returns results to the client.