Why Accuracy Can Be Misleading in Failure Detection Systems
When evaluating a machine learning model, accuracy is usually one of the first metrics we look at.
It is simple, intuitive, and easy to compare. But that simplicity can become a problem when the data
