Parallel probabilistic model-building genetic algorithms with elitism

This paper presents a parallel probabilistic model-building genetic algorithms (PMBGAs) called cellular compact genetic algorithm (CCGA) with elitism. The elitism-based CCCA is a coarse-grained parallel GA that migrates probability model between nodes instead of individuals. Each CCGA node is enhanced from compact genetic algorithm by using elitism. With elitism and our parallelized approach, the performance of the proposed parallel genetic algorithm is improved. The benchmarks and experimental results presented in the this paper confirm the performance of the proposed algorithm.