It is often the case that a simple term can have many different names. But terms used in Machine Learning are especially unique.

ROC curve: Receiver Operating Characteristic — sounds like something entirely new. But it’s essentially the same as the P-P plot in statistics. (HEP people often draw it mirrored.)

TPR (True Positive Rate): In machine learning, this is known as “signal efficiency” or just “s” in HEP terminology. In statistics, the survival function often conveys the same idea, though it’s rarely applied to distribution comparisons.

Log-likelihood: In statistics, it’s called logloss in machine learning. Neural network practitioners often refer to it as cross-entropy loss, while scikit-learn developers use the term binomial deviance. And there might still be other names I’m unaware of!

At first, these naming differences can drive you crazy. Over time, they blur into insignificance. But eventually, it becomes a real issue when the same concepts are discussed using different terms, leaving you unsure which one to use when searching online.