Time Series Prediction and Neural Networks
Neural Network approaches to time series prediction are briefly discussed, and the need to specify an appropriately sized input window identified. Relevant theoretical results from dynamic systems theory are introduced, and the number of false neighbours heuristic is described, as a means of finding the correct embedding dimension, and thence window size. The method is applied to three time series and the resulting generalisation performance of the trained feed-forward neural network predictors is analysed. It is shown that the heuristics can provide useful information in defining the appropriate network architecture. I. INTRODUCTION Neural Networks have been widely used as time series forecasters: most often these are feed-forward networks which employ a sliding window over the input sequence. Typical examples of this approach are market predictions, meteorological and network traffic forecasting. [1,2,3]. Two important issues must be addressed in such systems: the frequency with whi...