Overfitting remains a persistent challenge in deep learning, primarily attributed to data outliers, noise, and limited training set sizes. This thesis presents Divide2Conquer (D2C), a novel technique designed to address ...