Skip to Main Content
Neural network ensemble is a recently developed technology, which trains a few of neural networks and then combines their prediction results. The generalization ability of an ensemble can be much better than that of a single learner. However, it is time-consuming to train so many neural networks. In order to overcome this disadvantage, MPI (Message Passing Interface) technique was employed to the Parallel Implementation of the RBF neural network ensemble, which is named PDNNE in our previous work. In this paper, a unified parallelization strategy based on PDNNE is proposed to reduce the complexity and improve the efficiency of the ensemblepsilas construction for the bagging-based neural network ensembles (UPDNNE). And then a neural network ensemble learning platform (NNELP) based on the Shanghai University E-Grid Platform is designed and implemented to make the user conveniently and efficiently apply the ensemble methods to solve the real-world problems furthermore. The experiments on UCI datasets and the time sequence forecasting of strong earthquakes in Chinapsilas mainland demonstrate the efficiency of the NNELP.