Automating with AWS CloudFormation

In this tutorial, you use an AWS CloudFormation automation stack to launch an Amazon Forecast pipeline and generate forecasts using a demonstration dataset.

The AWS Forecast AWS CloudFormation stack:

• Deploys the Improving Forecast Accuracy with Machine Learning Solution AWS CloudFormation template.

• Deploys the NYC Taxi Datasets to the Forecast Data Amazon S3 bucket.

• Automatically starts the demo NYC taxi forecast pipeline in Forecast.

The AWS CloudFormation template is preloaded with target time-series, related time-series, and item metadata demonstration datasets. Relevant fields in the console are pre-filled with their respective S3 locations.

After completing this tutorial using the demonstration datasets, you can use the same automation stack to generate forecasts with your own datasets.

The following diagram shows the components used in this tutorial.

Prerequisites

Before starting the tutorial, make sure you have logged into your AWS account and installed the AWS CloudFormation template:

1. Log in to your AWS account. If you do not already have one, create an AWS account.

2. Install the AWS CloudFormation template. Choose the Region closest to you:

   • Tokyo: ap-northeast-1

   • Seoul: ap-northeast-2

   • Mumbai: ap-south-1

   • Singapore: ap-southeast-1

   • Sydney: ap-southeast-2

   • Frankfurt: eu-cental-1

   • Ireland: eu-west-1

   • N. Virginia: us-east-1

   • Ohio: us-east-2

   • Oregon: us-west-2

This deploys a demonstration stack using the NYC Taxi Dataset.

Deploying an AWS CloudFormation Template for Forecast automation

To deploy the CloudFormation template using the NYC Taxi Dataset

Step 1: Accept the defaults and choose Next.

Step 2: Provide an email address for notifications and choose Next.

Step 3: Accept defaults and choose Next.

Step 4: For Capabilities, select both check boxes to allow AWS CloudFormation to create AWS Identity and Access Management (IAM) resources and nested stacks. Choose Create stack.

You have deployed an AWS CloudFormation template in Forecast.

Clean Up

After deploying this AWS CloudFormation template, you can clean up newly created resources, deploy the AWS CloudFormation stack using your own datasets, and explore other deployment options.

• Cleaning up: Deleting the demo stack retains the "Improving Forecast Accuracy with Machine Learning" stack. Deleting the "Improving Forecast Accuracy with Machine Learning" stack retains all S3, Athena, QuickSight, and Forecast data.

• Using your own datasets: To deploy this AWS CloudFormation template with your own time-series data, enter the S3 locations of your datasets in the Datasets Configuration section in Step 2.

• Other deployment options: For more deployment options, see Automated Deployment. If data is already available, you can deploy the stack without the demo data.