AWS Solutions Implementation - The Improving Forecast Accuracy with Machine Learning solution adds contextual featurizations and visualization functionalities to enrich forecast insight

The AWS Solutions team updated Improving Forecast Accuracy with Machine Learning, an AWS Solutions Implementation that helps customers easily generate, test, compare, and iterate on forecasts. The solution allows customers to automatically trigger a workflow to automate the creation of forecasts in Amazon Forecast based on user-provided configuration. The solution then provides automatic visualization using Amazon QuickSight (and optionally Amazon SageMaker Jupyter Notebooks) providing a quick, easy, drag-and-drop interface that displays time series input and forecasted output for any input time series data.

This update allows customers to improve the accuracy of their forecast by leveraging the Amazon Forecast Weather Index and holidays featurizations. These automatic featurizations incorporate historical and projected weather information and national holiday information into their model to improve accuracy. To get started using the weather index and holidays featurizations, the solution also provides a comprehensive end-to-end demo of the solution using publicly available data.

Amazon QuickSight Enterprise customers can use the updated solution to visualize their model performance over their existing data, superimposing model forecasts over real data. This helps better visualize model performance alongside real-world data.

• Learn more about the Improving Forecast Accuracy with Machine Learning solution by visiting the AWS Solutions Implementation webpage.
• Watch an overview of the solution on Solving with AWS Solutions.

Additional AWS Solutions are available on the AWS Solutions Implementation webpage, where customers can browse solutions by product category or industry to find AWS-vetted, automated, turnkey reference implementations that address specific business needs.