The online social network has become an integral part of modern society and serves as an excellent platform for information diffusion, marketing, and so on. The minimum positive influence dominating set (MPIDS) problem aims to find a set of at least half neighbors of every individual with minimum cardinality and has lots of applications in the social network. Even though numerous algorithms have been proposed for this hard combinatorial problem, challenges such as poor quality and scalability have motivated the researchers for better solutions. In this article, we focus on designing a simple and effective method and propose an iterated carousel greedy (ICG) algorithm for the MPIDS problem. The destruction, carousel, and reconstruction (DCR) procedures are well-designed based on the problem-specific knowledge to improve the algorithm performance, while maintaining the simplicity of the ICG algorithm. The ICG algorithm uses a fast greedy algorithm to generate an initial solution with a certain quality. It then obtains a sequence of better solutions by iterating over DCR phases. A carefully designed experiment is carried out to determine the best parameter configuration. The experimental results on real-world networks indicate that the proposed ICG algorithm significantly outperforms the current state-of-the-art algorithms.

Task scheduling is one of the core steps to effectively exploit the capabilities of distributed heterogeneous computing systems. In this paper, a novel discrete differential evolution (DDE) algorithm is presented to address the task scheduling problem. The encoding schemes and the adaptation of classical differential evolution algorithm for dealing with discrete variables are discussed as well as the technique needed to handle boundary constraints. The performance of the proposed DDE algorithm is showed by comparing it with a genetic algorithm, which is a well-known population-based probabilistic heuristic, on a large number of randomly generated instances. Experimental results indicate that the proposed DDE algorithm has generated better results than GA in terms of both solution quality and computational time.