Summary

In this chapter, we saw how to use Redshift as a datasource for Amazon ML. Although RDS could also have been used to create datasources, Redshift is much easier to use with Amazon ML as all the access configuration is taken care of by the AWS wizard. 

We have shown how to use simple SQL queries on Redshift to carry out feature engineering and implement a polynomial regression approach on a highly non-linear dataset. We have also shown how to generate the required SQL queries, schemas, and recipes to carry out the Monte Carlo cross-validation.

In the next chapter, we will build on our Redshift integration and start streaming data using the AWS Kinesis service.