Time series forecasting using multilayer neural network constructed by a Monte-Carlo based algorithm
Peiming Xin, Hong Wang Zhao · 2009
In this paper we use a multilayer neural network (MNN) constructed by a Monte Carlo based algorithm to forecast time series events. Experiments are carried out on two benchmark problems in time series forecasting literature. Our result shows a comparative performance with other prediction methods. In our approach, a multilayer neural network with high level of generalization ability is obtained without sensible choice of external parameters.