AWS re:Invent 2019: End-to-end machine learning using Spark and Amazon SageMaker (ADM302-R1)

Learn how AWS customers are developing production-ready machine learning models to optimize auction dynamics and bid pricing—all within the millisecond latency requirements of programmatic ad buying. Forget “hello world” ML tutorials; instead we dive deep into an example of how to train models for terabyte-scale advertising data cost-effectively. Find out how to create environments for machine learning engineers so they can prototype and explore with TensorFlow before executing it in distributed systems using Spark and Amazon SageMaker. We explain productionizing models and deployment, which we then back up with real-life examples.