Robust Sequence Networked Submodular Maximization

In this paper, we study the Robust optimization for sequence Networked submodular maximization (RoseNets) problem. We interweave robust with networked maximization. The elements are connected by a directed acyclic graph and objective function is not on but edges in graph. Under such scenario, impact of removing an element from depends both its position network. This makes existing algorithms inapplicable calls new algorithms. take first step to RoseNets design greedy algorithms, which against removal arbitrary subset selected elements. approximation ratio algorithm number removed network topology. further conduct experiments real applications recommendation link prediction. experimental results demonstrate effectiveness proposed algorithm.