This example notebook shows how to use mutiple models from SageMaker for prediction and combine then into an ensemble prediction.
It demonstrates the following:
- Basic setup for using SageMaker.
- converting datasets to protobuf format used by the Amazon SageMaker algorithms and uploading to user provided S3 bucket.
- Training SageMaker's XGBoost algorithm on the data set.
- Training SageMaker's Linear Learner on the data set.
- Hosting the trained models.
- Scoring using the trained models.
- Combining predictions from the trained models in an ensemble.