Rahmalia Nurida, - and Endang Sugiharti, - and Alamsyah, FMIPA Ilkom (2019) Implementation of Fuzzy K-Nearest Neighbor Method in Decision Support System for Identification of Under-five Children Nutritional Status Based on Anthropometry Index. Journal of Advances in Information Systems and Technology 1, 1 (1). pp. 83-89. ISSN 2714-9714

	PDF - Published Version Download (2MB)
	PDF - Published Version Download (1MB)

Abstract

Nutritional status is one of the important factors in assessing the level of health and growth of infants and under-five children. But the present, there are still many problems caused by an imbalance in nutritional intake with the nutritional needs of children. K-Nearest Neighbor method in the previous studies showed the existence of prediction results with the problem. This study used a standard anthropometric index or body size to carry out the process of calculating nutritional status using the Fuzzy K-Nearest Neighbor method. Fuzzy is applied to reduce the problem in classification. Predictions produced are three categories of nutritional status, namely BB/U (weight according to age), TB/U (height according to age), and BB/TB (weight according to height). The k value taken for the classification process is k=10. Fuzzy K-Nearest Neighbor process has done by taking the closest Euclidean distance to the number k from the training data to the test data. The prediction class results of 95 out of 96 data are stated accordingly after the value of the membership is calculated. The accuracy of the test performed produces a cumulative accuracy of 98.96%. This study can be used as a reference for further research by adding training data with more complete class variations in each category of nutritional status to obtain more optimal accuracy.

Actions (login required)

View Item