In the great majority of human intellectual activities, knowledge acquisition involves several levels of imprecision and uncertainty. This is particularly true in epidemiological and medical diagnosis problems. Recently, there has been an increase of the interest of epidemiologists, physicians and ecologists in fuzzy theory, with the aim of treating their problems in a more realistic way. Modelling epidemiological systems is a difficult task. This is particularly true for fuzzy epidemics models which usually rely heavily on experts knowledge. Our own experience in dealing with epidemiology and fuzzy epidemic modelling have showed us that in these case of biological, and particularly epidemiological, models it is often hard to find functional information shout the dynamics of the systems. In this context, the academic and heuristic knowledge of experts, as well as their experience, assume a fundamental role in this kind of modelling. However, experts may have serious problems to insight both the antecedents and the consequents of the rules when the model is oomplex. Furthermore, to create the consequents is a much more difficult task than the antecedents because in the former the expert needs to consider the dynamics of the system, weighting all influences that could concur, generating one specific output and its corresponding membership function. On the order hand, in order to create the antecedents, the expert needs only to classify the variables in groups, elaborating their membership functions. Therefore, in general, the expert has more facility to elaborate the antecedents than the consequents. In this sense, a method that allows the elaboration of the consequents of the linguistic rules would imply in an important progress in the modeling of systems which have a high level of uncertainties, impreciseness and/or vagueness in the variables, parameters or both. In addition,
展开▼