Skip to main content

Bias detection and explainability in AI and machine learning applications

AI services and machine learning are helping organisations to build data driven applications that are innovative and can be highly attuned to their customers’ needs, but AI applications require crucial customer data to train machine learning models. Application logic is delegated to these models, which can introduce unfairness and biases into an application. In this session we walk through the machine learning techniques and AWS services you can use to understand and reduce these risks.