TY - RPRT TI - Adaptive norms for deep learning with regularized Newton methods AU - Jonas Kohler AU - Leonard Adolphs AU - Aurelien Lucchi PY - 2020 UR - https://arxiv.org/abs/1905.09201 ID - 1905.09201 ER -