A fuzzy expert system for diagnosis of malnutrition in children

The aim of this study is to design a fuzzy expert system which diagnoses whether an infant is suffering from malnutrition or not, and if he/she is in the trap of malnutrition, then what is the extent or severity of the same. For this purpose, we have taken into account 13 input variables-breast feeding, daily family income, level of mother's education, colostrum intake, energy intake, protein intake, vitamin A intake, iron intake, family size, height of infant, weight of infant, head circumference of infant and skinfold thickness of infant. The presence of malnutrition and its severity are diagnosed, keeping the following categories in view-severe under nutrition, moderate under nutrition, mild under nutrition, normally nourished, mild over nutrition and moderate over nutrition. At this very point, it is worth stressing the fact that malnutrition is not just an indicator of under nutrition. Over nutrition is as well a form of malnutrition. The designed expert system uses Mamdani inference method. With this system, one can very easily diagnose the level of nourishment in an infant by just giving appropriate values of the inputs.