Klasifikasi Status Gizi Balita Berdasarkan Indikator Antropometri Berat Badan Menurut Umur Menggunakan Learning Vector Quantization

Abstract

Determination of nutritional status is an effort made in order to improve the health of children. Common method used for the assessment of nutritional status is anthropometry. To classify the nutritional status of children into malnutrition, malnutrition, good nutrition and nutrition then used anthropometric indices weight for age (W / A). In Rimbo data Puskesmas, calculation of anthropometric indices for the assessment of nutritional status of children is done manually using z-scores table lists or standard deviation (SD) WHO NCHS. In this research, the authors tried to establish a classification system based nutritional anthropometric indices weight for age (W / A) by applying the Learning Vector Quantization algorithm uses two functions, namely euclidean and manhattan distance. The variables used were gender, age, weight, family economic status, mother's education, father's occupation. From the results of research and discussion conducted, Learning Vector Quantization algorithm using euclidean distance function can recognize the pattern with the best accuracy percentage of 80% whereas the manhattan distance function only 20% of 110 training data and test data amounted to 10. The amount of training data and the diversity of patterns that exist in the class used nutritional status affects learning outcomes and the accuracy of the systems
Keywords: Antropometri, Euclidean, Learning Vektor Quantization, Manhattan, Z-skor