The aim of the paper is to demonstrate - by means of two exemplary approaches - how modelling can help to analyse and understand the present state as well as predict the future behaviour of a dynamic system. Therefore in the following first of all some examples are discussed, focusing on basic aspects of system dynamics and prediction. Chapter 2 then deals with an application to performance analysis, which in Chapter 3 is transferred to neural network-based analysis of adaptive processes like learning. Modelling a dynamic system means mapping not only its components and input-output-behaviour but - depending on the demand on precision - also and in particular its components interaction. The reason is that dynamic interaction of system components is characterized by buffers and delays. This means that a system input -depending on the respective internal status - can cause quite different outputs, which rnakes it extrernely difficult to predict the system behaviour by only taking static pairs of input- and output-values into consideration. Accordingly, stochastic analysis methods like correlation or trend analysis based on past values in case of dynamic Systems are normally neither sufficient nor adequate for an acceptable prediction.
© Copyright 2004 International Journal of Computer Science in Sport. Sciendo. Alle Rechte vorbehalten. / Übersetzt mit https://www.DeepL.com/Translator