AIDIL ZIKRI, -
(2024)
THE RANDOM FOREST ALGORITHM FOR CLASSIFYING STUNTING IN TODDLERS BASED ON ANTROPOMETRIC DATA.
International Journal of Multidisciplinary Research and Growth Evaluation.
ISSN 2582-7138
Abstract
Stunting is a growth and developmental disorder in children that often occurs during the first 1,000 days of life, from conception to the age of two. Factors such as inadequate nutritional intake, recurrent infections, and a less clean environment can contribute to stunting. In this case, use the random forest algorithm. The goal is to categorize the stunting case. The variables used are gender, age, birth weight, birth height, weight, height, and Exclusive Breastfeeding. With the number of datasets used reaching 6,500, the experiment was performed with a combination of parameters, namely n_estimators 100, 200, and 500. Max_features = 5, 7, 10, and 13. As well as min_samples_split and min_sample_leaf = 20, 50, and 100. Based on the specified set of hyperparameters, 108 test results were obtained. The highest accuracy is 0.9651, with precision = 0.9603, recall = 0.9718, and F1-score = 0.9660. With n_estimators formed = 100, max_depth = 13, min_samples_split = 20, and min_samples_leaf = 20.
Actions (login required)
 |
View Item |
