<![CDATA[Malnutrition among children under five remains a critical public health challenge, particularly in primary healthcare settings where assessment is often conducted manually and relies on a single anthropometric index. This study proposes a Mamdani fuzzy logic-based classification system designed to assess children’s nutritional status at Puskesmas Nanu by simultaneously incorporating four anthropometric parameters: age (months), height, weight, and mid-upper arm circumference (MUAC). Unlike previous studies that typically employ one or two indicators, this system constructs a comprehensive inference framework consisting of 135 IF-THEN rules derived from all possible combinations of fuzzy input sets. Triangular and trapezoidal membership functions are applied to each variable to capture the gradual transitions inherent in children’s growth conditions. The inference engine employs the MIN operator for rule activation and MAX for aggregation, while centroid defuzzification converts the aggregated fuzzy output into a deterministic crisp value. The system was evaluated against 20 anthropometric records from the facility and compared with the conventional Z-score method used by healthcare workers. Results show that 15 out of 20 cases were classified consistently, yielding an accuracy rate of 75%. In a representative case of a 59-month-old child, the system produced a crisp output of −0.25, corresponding to the normal nutritional status category. These findings demonstrate that the proposed system offers a more holistic and objective approach to nutritional assessment. Limitations include the relatively small sample size and membership function domains derived from local data rather than standardized WHO references. Future work should focus on expanding the dataset, aligning parameters with national anthropometric standards, and implementing the system as a web-based or mobile application integrated into primary healthcare information systems.]]>
Copyrights © 2026
