A/B testing machine learning models with Amazon SageMaker MLOps

April 22, 2021
Machine learning operations (MLOps) enable developers and scientists to improve their models for various reasons such as training on recent data, altering features, or trying new algorithms. Performing A/B testing between the new model and the old model will be an effective step in the validation process when deploying a new model into production. In this session, learn how to build an Amazon SageMaker MLOps pipeline to automate the training and deployment of multiple models for predicting product review helpfulness. We also share how a multi-armed bandit experiment framework can automatically optimise traffic to the best-performing model over time based on user feedback.
Previous Video
Using reinforcement learning to solve business problems
Using reinforcement learning to solve business problems

Join this session to learn how to structure your business problem into a reinforcement learning structure, ...

Next Video
Computer vision at the edge
Computer vision at the edge

Computer vision is used across most industries to help solve business challenges, from automotive, manufact...