Neural Network Approaches to Unimodal Surjective Map Chaotic System Forecasting

The forecasting using neural networks in unimodal surjective map chaotic dynamic system will be studied carefully in this paper. And most of the forecasting precision has exceeded 90%. Because of the intrinsic property of chaos, the forecasting precision will decrease as the length of symbolic sequence is increasing. But in this place we have found a generating rule that may realize chaotic synchronization at least in short and medium term, and we can analysis and forecast in this way. Nonlinear dynamics maintain manifold links with biologic information system. We also hope to offer an effective prediction method to study certain properties of DNA base sequences, 20 amino acids symbolic sequences of proteid structure, and the time series that can be symbolic in finance market et al.