Posted On: Oct 1, 2020
Amazon SageMaker Autopilot now creates machine learning models up to 40% faster with up to 200% higher accuracy, even with small and imbalanced datasets. SageMaker Autopilot is a fully managed service that makes it easy to quickly and automatically create machine learning models, without requiring any prior ML or programming experience.
Until today, SageMaker Autopilot provided the highest quality models for large and balanced datasets. Now, SageMaker Autopilot can provide models that are up to 200% more accurate even when datasets are highly imbalanced, and have as few as 500 data points. You can also now direct SageMaker Autopilot to use AUC, or the “area under the receiver operating characteristic (ROC) curve” objective metric to create even more accurate models for classification problems. Additionally, SageMaker Autopilot has increased the speed at which models are created by up to 40%, allowing you to begin extracting business value from your data faster. With these enhancements, you can easily use SageMaker Autopilot to quickly create and deploy high quality classification and regression machine learning models, make accurate predictions, and extract value from datasets of a wide range of sizes, quality, and data types.
The new performance improvements are available in all regions where SageMaker Autopilot is available. To get started with SageMaker Autopilot visit the documentation page and watch the getting started videos.