Open AccessModeling of toddler stunting in the province of east nusa tenggara using multivariate adaptive regression splines (mars) method
Author(s) -
D M Azizah,
Erma Oktania Permatasari
Publication year - 2020
Publication title -
journal of physics. conference series
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.21
H-Index - 85
eISSN - 1742-6596
pISSN - 1742-6588
DOI - 10.1088/1742-6596/1490/1/012013
Subject(s) - toddler , multivariate adaptive regression splines , malnutrition , mars exploration program , multivariate statistics , demography , medicine , regression analysis , geography , environmental health , pediatrics , statistics , bayesian multivariate linear regression , mathematics , psychology , biology , developmental psychology , pathology , sociology , astrobiology
Stunting is a condition of failure to thrive in children under five (babies under five years old) as a result of chronic malnutrition so that the child is too short for his age. Malnutrition occurs since the baby is in the womb and in the early period after the baby is born. However, the condition of stunting only appears after the baby is 2 years old. Based on Basic Health Research in 2017 the province with the highest prevalence of short and very short toddlers in Indonesia is East Nusa Tenggara which is equal to 40.3 percent. This shows that the prevalence of stunting toddlers in NTT Province is still far from the overall prevalence of stunting toddlers in Indonesia. Cases of toddler stunting can be analyzed using the Multivariate Adaptive Regression Splines (MARS) method because the data used are high-dimensional data and do not show clear patterns of relationships between response variables and predictor variables. The data used is the percentage of stunting toddlers in NTT in 2017 with the district / city research unit. Based on the analysis, it was found that the district in NTT with the highest percentage of stunting children was South East Timor and the district with the lowest percentage was East Manggarai. The best MARS model is formed from 3 basis functions with 2 predictor variables included in the model, namely the percentage of pregnant women at risk of chronic lack of energy (X1) and the percentage of infants with low birth weight (X6). And after the classification results obtained the accuracy of the classification of 77.27% with an error of 22.72%.