Working with more built-in algorithms

In the rest of this chapter, we will run more examples with built-in algorithms, both in supervised and unsupervised mode. This will help you become very familiar with the SageMaker SDK, and learn how to solve actual machine learning problems. The following list shows some of these algorithms:

• Classification with XGBoost
• Recommendation with Factorization Machines
• Dimensionality reduction with PCA
• Anomaly detection with Random Cut Forest

Classification with XGBoost

Let's train a model on the Boston Housing dataset with the XGBoost algorithm (https://github.com/dmlc/xgboost). As we will see in Chapter 7, Using Built-in Frameworks, SageMaker also supports XGBoost scripts:

1. We reuse the dataset preparation steps from the previous examples.
2. We find the name of the XGBoost container. As several versions are supported, we select the latest one (1.0-1 at the time of writing):

   import boto3 from sagemaker import...