Immunization of complex networks using stochastic hill-climbing algorithm

Recently, there is a growing interest in how to mitigate epidemic spreading through complex networks such as infection propagation in population, rumor spreading in social interaction, and, malicious attacks in computer networks. Due to high cost and limitation of immunization resources, a well-established strategy is required to select whom to inoculate. In this paper, we propose a new immunization strategy based on stochastic hill-climbing algorithm to find a subset of nodes whose immunization efficiently reduce the network vulnerability to worst-case epidemic size. Our experiments show that SHCI shows up to 31% improvement in real networks and up to 89% in model networks compared to targeted immunization algorithms which immunize nodes based on their centrality.