Distributed Constraint Programming based on Fuzzy Probability Applied to Aphasia Diagnosis

Fuzzy constraint programming is implemented to optimize a cost function facing vague constraints. In this paper, for the first time, a novel distributed constraint programming is developed by taking into account fuzzy and probability concepts. In the proposed approach, we use the human way of thinking by expressing constraints in the form of rules. The proposed approach can deal with different kinds of uncertainty. We model the content of constraints using not only fuzzy inequality constraints but also probability inequality. Then, a proposed distributed protocol is applied based on the cooperative constraints programming regulation framework An adaptive certainty is adjusted for the set of constraints depending on the relevant satisfaction degree. Then, the problem of constraint programming is converted to the optimization problem. This approach can be considered a step forward for modeling complex systems. The proposed approach has been explored by the application of aphasia diagnosis facing inconsistency in the clarification of aphasic syndromes. Even by requiring more executed time, the improvement of results’ accuracy is observed especially when compared to the results of alternative approaches applying the same database.