Discrete Process Neural Networks and Its Application in Time Series Prediction

Considering that inputs of a process neural network (PNN) are generally time-varying functions while the inputs of many practical problems are discrete values of multiple series, in this paper, a process neural network with discrete inputs is presented to provide improved forecasting results for solving the complex time series prediction. The proposed model first makes the discrete input series carry out Walsh transformation, and then submits the transformed series to the network for training, which can solve the problem of space-time aggregation operation of PNN. In order to examine the effectiveness of the proposed method, the two examples are employed. First, the developed model is tested on the Mackey-Glass time series and has comparison with the results in literatures, and then, taking the actual data of sunspots during 1749-2007 as examples, the number of sunspots is predicted and the suitability of the developed method is examined in comparison with the other models to show its superiority. The proposed method provides a new way for the space environment prediction in future. KeywordsProcess neuron; Process neural networks; Learning algorithm; Time series predication; sunspot number