Developing an influence diagram using a Structural Modeling, Inference, and Learning Engine

In this paper, we proposed a method for developing Influence Diagrams by using the engine namely, Structural Modeling, Inference, and Learning Engine (SMILE). An influence diagram is a simple visual representation of a decision problem. It can be used to assist in rational decision making for different purposes. We proposed a practical method and a piece of source code in C#.net to build an influence diagram based on a SMILE engine. We presented the steps on how to set four types of nodes and their properties in an influence diagram using C#.net. This piece of source code also includes the method to construct each component such as decision, chance, deterministic, value and arcs between the nodes in the diagram. It also includes the method to perform Bayesian updating and Bayesian inference according to Baye's theorem. The developer can apply this proposed method for the extension model of influence diagrams. He or she can also embed this piece of source code in an existing decision support application.