Soft Computational (Neural Nets) Dyadic Interface to System Dynamics in Developing Business Expert System

Application of systems science in business has lately emerged as an exciting area of applied research. It aims to study the behavior of complex business systems. The present study attempts to converge the two powerful approaches of system approximation in form of artificial neural network and system dynamics which are based on rigorous modeling and simulation processes. It also envisages a dynamic possibility of data and decision interface generated between the two in order to improve the overall efficiency of a representational system design and decision making.As a consequence of this interface, neural network can be used for enriching the SD model in a way that it would assist a SD modeler in policy stabilization of input variables by maintaining a value of target variables within a desired level. Here, the luxury available to the experimenter is enormous amount of data generated through SD simulation which would be used for training and testing the neural network