Phase Space Reconstruction and Time Series Prediction of a Nonlinear Financial System

As a mechanism for trading financial assets and determining the price of financial assets, the financial market plays an important guiding role in the development of the social economy. To explore the essential characteristics of the financial market and analyze its potential development laws, this paper takes a classical financial system as an example to reconstruct the chaotic attractor of the system based on a one-dimensional series. We adopt the C-C method to calculate the optimal delay time and embedding dimension that are two important indicators used to reconstruct the chaotic attractor. Furthermore, the BP neural network is used to predict the financial market trend. We discuss the predicted outputs for the BP neural network with different numbers of neurons or hidden layers. Numerical results show that the proposed BP neural network is effective to predict market trends based on the existing data.