Universality of High-Dimensional Logistic Regression and a Novel CGMT under Block Dependence with Applications to Data Augmentation

Open in new window