Three qualitative simulation extensions for supporting economics models

Three extensions to qualitative simulation (QSIM) that permit modeling and simulation of a large class of economics problems are described. To be more powerful, financial expert systems need qualitative causal models in addition to associative knowledge. QSIM is a useful tool for building such models. However, building the models requires some extensions to QSIM. In particular, QSIM is unable to simulate the combination of average and marginal parameters or the constant time delays between actions and their effects, and it cannot treat certain statistics problems dynamically. Ways of overcoming these limitations so that a larger class of economics models can be simulated qualitatively are identified, and their implementation is described.<
>