Logic, Reasoning and A Programming Language for Simulating Economic and Business Processes with Artificially Intelligent Agents

9 A Modelling Methodology using SDML

The formal tools of logical modelling and the computer tool of SDML allow the modeller to go beyond the limitations of statistical or differential equations and begin to tackle the complexity inherent in many of the management and business situations faced today.

A possible outline for the use of such tools might be the following.

Choose a target for modelling for which a formal model of its separate causal parts is credible.
The intuitions of the causes of the agents' actions are encoded into the rules governing their "actions" in the formal system (carefully remembering any assumptions).
These rules are then animated by the deductive machinery of SDML.
The results are compared to the known data on the effects of this process, according to previously decided criteria.
If the results do not match the data according to these criteria, then go back to step (2). Ensure that if a loop of steps (2)-(5) occurs too many times then you more fundamentally change the model (or give up on this process).
If the results are successful examine the emergent process of the animation by the formal system, interpreting it into qualitative conclusions (minding your assumptions).
If appropriate make hypotheses about the model and attempt to prove them.
Recursively apply any insights to further modelling of this kind, to more traditional types of models or other informal decision making.

This process is illustrated in figure 3 below.

Figure 3: A Methodology for simulation

Thus we can continue to make substantial progress in domains where traditional analytic modelling tools are impractical.

Logic, Reasoning and A Programming Language for Simulating Economic and Business Processes with Artificially Intelligent Agents - 12 APR 96

[Next] [Previous] [Top] [Contents]

Generated with CERN WebMaker