A fuzzy neuro approach to identify diarrhea epidemic in Bangladesh
Muhammad Swadhin Shahriar Faruque, Shourav Banik, Rashedur Mohammad Rahman · 2014
The paper presents the identification of probability of diarrhea epidemic occurring in Bangladesh per month using a competitive neural network and fuzzy logic. Here we have divided the months into six seasons: spring, summer, rainy, early fall, late fall, winter. The infected rate is divided into four parts: low, medium, high, very high. At first infection rate in each season is learned by using a competitive neural network and then the identification of the percentage of an epidemic occurrence is done by fuzzy algorithm (specifically by the Mamdani Min). The centroid function was later used to get a crisp value that corresponds to the probability of epidemic in a certain year.