Multi-step prediction of pulmonary infection with the use of evolutionary fuzzy cognitive maps

The task of prediction in the medical domain is a very complex one, considering the level of vagueness and uncertainty management. The main objective of the presented research is the multi-step prediction of state of pulmonary infection with the use of a predictive model learnt on the basis of changing with time data. The contribution of this paper is twofold. In the application domain, in order to predict the state of pneumonia, the approach of fuzzy cognitive maps (FCMs) is proposed as an easy of use, interpretable, and flexible predictive model. In the theoretical part, addressing the requirements of the medical problem, a multi-step enhancement of the evolutionary algorithm applied to learn the FCM was introduced. The advantage of using our method was justified theoretically and then verified experimentally. The results of our investigation seem to be encouraging, presenting the advantage of using the proposed multi-step prediction approach.