Abstract: This correspondence discusses a space–time adaptive processing (STAP) method using deep neural network (DNN) and shrinkage algorithm (SA) under small training samples. A DNN-based SA is ...