Intuition, algorithm and code for using ALEs to explain machine learning models
Highly correlated features can wreak havoc on your model interpretations. They violate the assumptions of many XAI methods and make it difficult to understand the nature of a feature’s relationship with the target. At the same…