The Time Series Prediction Based On Echo State Network Of Modified Small World

Time series exist widely in industry, economy, military affairs and other fields, such as blast furnace gas output, the Shanghai composite index of stock market, mobile telephone traffic, the change of temperature. There are many historical information in the time series. In order to analyze the future dynamic of time series, we make hypothesis and inference by using the historical information and their correlation. This process is time series forecasting. Time series forecasting has always been researched in the various fields widely. It is researched that the time series prediction based on echo state network of modified small world in this paper.In order to improve the prediction accuracy of echo state network for time series, this paper proposes a modified small world network to improve Leaky-integrator echo state network (Leaky ESN) and utilizes the improved Leaky ESN to predict time series. Firstly, we propose a modified small world network whose adding edge probability is a negative exponential function of the distance between the nodes. Then, we utilize adding edge probability to directly represent the connection weights of two neural nodes in reservoir of Leaky ESN. And the value of the connection weights belongs to [0,1] and represents the degree of connection between the two nodes. The modified small world network is utilized to improve the connection of reservoir neural nodes, which purposefully achieve the sparse connection to reduce the blindness of the original sparse connections and then improve the adaptive capacity of reservoir. Finally, we use the modified Leaky ESN to predict a typical nonlinear time series. We use Matlab software to verify the effectiveness of the method proposed in this paper. Compared with Leaky ESN, the method proposed in this paper has a higher prediction precision and a shorter training time.