Towards predicting neural net control of macro-econometric multi-compartment models

We propose a novel concept for analyzing dynamic economic and financial systems by multi-compartment modeling techniques in combination with neural networks. This is done by a two-step process: 1) a multilayer perceptron (MLP_ident) is introduced to approximate the dynamics of observable variables of a single (domestic) compartment and to identify interdependencies; and 2) a multilayer perceptron (MLP_control) evaluates series of outputs from MLP _ident to make a complex decision about certain properties (e.g. stability) and to find a classification of economic or financial scenarios. Simulation results confirm the ability of MLP_ident to learn the structure and parameters of the domestic compartment and to provide good results in out-of-sample tests. Simulation results for MLP _control show that it is able to decide whether a current parameter set leads to a stable or unstable constellation in the future