Ponencias

Examinar

Mostrando 1 - 1 de 1

Integer programming applied to rule based systems
(Elsevier (Países Bajos), 2012) Ávila Herrera, Juan Félix
Abstract. In this paper we show how to represent a set of logic propositions as an integer linear program and how to use its solution to determine the truth value of all the proposition given only a subset of their truth values. Thus a rule–based expert system might use this approach as inference engine. When we model an expert system knowledge base as an integer linear program, we can ask for what is a minimum set of premises needed in order to have an specific conclusion as true. Furthermore we can ask for what propositions become true once a subset of propositions are known to be true. Once the integer restriction is dismissed, we can ask for what conditions are necessary in order that a conclusion holds with certain probability. Thus we can have both a deterministic and stochastic model using practically the same paradigm.