Amazon SageMaker Canvas is a visual, no-code machine learning service that enables business analysts and domain experts to build accurate ML models and generate predictions. This solution guides users through the complete workflow of preparing data and building a machine learning model using Amazon SageMaker Canvas, with direct integration to their Snowflake data warehouse. The integration reduces data movement challenges, allowing users to connect directly to their Snowflake data warehouses, transform the data within Canvas, and proceed straight to model building. This unified, no-code environment accelerates time-to-insight while maintaining data governance and security.
To begin, users connect Amazon SageMaker Canvas to a Snowflake data source by navigating to the Data Wrangler section and specifying the Snowflake connection type. They provide the necessary credentials, including the connection name, account ID, username, and password. Once the connection is established, they create card-level outlier thresholds to identify unusual spending patterns for each credit card and category combination. This helps the model detect when a transaction amount significantly deviates from a cardholder’s typical behavior. The SQL query to prepare the fraud detection dataset is then executed, and the results are imported into the Canvas workspace.
The process includes data enrichment with transaction-level features and merchant-specific fraud indicators. Users add data sources and run SQL queries to extract temporal patterns, demographic information, and merchant-level outlier thresholds. These features are critical for fraud detection. The data is then joined, transformed, and outliers are identified, with sensitive columns removed. The final dataset is ready for model building using the XGBoost algorithm.
Source: awsml
