A new time series prediction method based on complex network theory

In this work, we propose a new time series prediction method based on complex network theory, named Data Fluctuation Networks Predictive Model (DFNPM). The basic idea of the method is: to first map the time series into data networks and extract fluctuation features of time series accordingly to the topological structure of the data networks, and then construct models with useful information extracted to predict time series. We compare our model with the traditional prediction models as Grey Prediction Model (GM), Exponential Smoothing Model (ESM), Autoregressive Integrated Moving Average Model (ARIMA) and Radial Basis Function Neural Network (RBF) using crude oil and gasoline future prices. We obtained that the accuracy of DFNPM is higher than that previously cited models.