Regularization-based pruning of irrelevant weights in deep neural architectures
Sign Up to like & getrecommendations! Published in 2022 at "Applied Intelligence"
DOI: 10.48550/arxiv.2204.04977
Abstract: Deep neural networks exploiting million parameters are currently the norm. This is a potential issue because of the great number of computations needed for training, and the possible loss of generalization performance of overparameterized networks.… read more here.
Keywords: pruning irrelevant; deep neural; regularization; based pruning ... See more keywords