Detect Potential Bias in your Datasets and Explain how your Models Predict using SageMaker Clarify (Level 300)

June 9, 2021
As ML models are built by training algorithms that learn statistical patterns present in datasets, several questions immediately come to mind. First, can we ever hope to explain why our ML model comes up with a particular prediction? Second, what if our dataset doesn’t faithfully describe the real-life problem we were trying to model? Could we even detect such issues? Would they introduce some sort of bias in imperceptible ways? These are not speculative questions at all, and their implications can be far-reaching. Unfortunately, even with the best of intentions, bias issues may exist in datasets and be introduced into models with business, ethical, and legal consequences. It is thus important for model builders and administrators to be aware of potential sources of bias in production systems. In addition, many companies and organizations need ML models to be explainable before they can be used in production. In fact, some regulations explicitly require model explainability for consequential decision making. In this hands-on session, you’ll learn how Amazon SageMaker Clarify can help you tackle bias and explainability issues, and how to use it with both the SageMaker Studio user interface and the SageMaker SDK. You’ll also see how it works together with SageMaker Model Monitor to track bias metrics over time on your prediction endpoints.

Speaker: Julien Simon, AWS Principal Advocate, ML/AI
Previous Video
Scale your Large Distributed Training Jobs with Data and Model Parallelism Optimized for Amazon SageMaker [Level 300]
Scale your Large Distributed Training Jobs with Data and Model Parallelism Optimized for Amazon SageMaker [Level 300]

In this session, explore how to choose the proper instance for ML training and inference based on model siz...

Next Video
Select the Right ML Instance for your Training and Inference Job (Level 300)
Select the Right ML Instance for your Training and Inference Job (Level 300)