Extensions of graph inversion to support an artificially intelligent modeling environment

This paper extends previous results in graph inversion to enable artificially intelligent model formulation by taking partial information from the problem domain into an algebraic form-that is, a mathematical model to support economic based decision-making. Here focus is on flow relations and economic correlations. Issues of consistency are resolved and a concept of core models is introduced to achieve minimality for simplifying model management and use. After establishing the general results, special consideration is given to input-output structures, as in Leontief models.