Implementing linear regression through SageMaker's Linear Learner

Another alternative within AWS for training regression models is to use SageMaker's API to build linear models. In the previous chapter, we explained the basics of this service when we considered how to use BlazingText for our text classification problem. Similarly, we will use Linear Learners in this section and go through the same process, which basically entails three steps:

1. Stage the training and testing data in S3
2. Invoke the API to train the model
3. Use the model to obtain predictions

Unlike what we did in Chapter 2, Classifying Twitter Feeds with Naive Bayes, instead of deploying an endpoint (that is, a web service) to obtain predictions, we will use a batch transformer, which is a service that's capable of obtaining bulk predictions given a model and some data in S3. Let's take a look...