1 paper
Mou Wu, Haibin Liao, Zhengtao Ding +1
Gradient-type distributed optimization methods have blossomed into one of the most important tools for solving a minimization learning task over a networked agent system. However,…