Methoden der Datenvorverarbeitung und ihre Auswirkungen auf ANNs

In the analysis using artificial neural networks (ANN), noisy, non-stationary data is often considered for time series forecasting. This book addresses various data preprocessing methods and how they can enhance the prediction accuracy of artificial neural networks. Choosing the right configuration for an ANN is a non-trivial task, and the results obtained are verified through the application of different criteria and the exchange of experimental and test series, as well as through randomization, ensuring their applicability even under extreme conditions.