Time Series Prediction with Dynamic Neural Networks - MATLAB-Based NARX Implementation

Time series forecasting holds significant importance in economics and engineering fields. This study leverages the characteristics of dynamic neural networks to propose a time series prediction methodology, implementing a designed dynamic network to forecast response time series of Duffing's equation. Results demonstrate that dynamic neural networks effectively predict response time series of dynamic systems, with MATLAB implementation utilizing NARX (Nonlinear Autoregressive with Exogenous Input) network architecture and time-delay feedback mechanisms.

Neural Network Prediction

Dynamic Neural Networks - Adaptive Architectures and Implementation Approaches