Universal Neyman-Pearson Classification with a Known Hypothesis

Parham Boroumand, Albert Guillén i Fàbregas

We propose a universal classifier for binary Neyman-Pearson classification where null distribution is known while only a training sequence is available for the alternative distribu…