Asynchronous Distributed Parallel Gene Expression Programming Based on Estimation of Distribution Algorithm

In order to reduce the computation time and improve the quality of solutions of Gene Expression Programming (GEP), a new asynchronous distributed parallel gene expression programming based on Estimation of Distribution Algorithm (EDA) is proposed. The idea of introducing EDA into GEP is to accelerate the convergence speed. Moreover, the improved GEP is implemented by an asynchronous distributed parallel method based on the island parallel model on a message passing interface (MPI) environment. Some experiments are done on distributed network connected by twenty computers. The best results of sequential and parallel algorithms are compared, speedup and performance influence of some important parallel control parameters to this parallel algorithm are discussed. The experimental results show that it may approach linear speedup and has better ability to find optimal solution and higher stability than sequential algorithm.