Naura, Sheila Sevira Asteriska
Unknown Affiliation

Published : 2 Documents Claim Missing Document
Claim Missing Document
Check
Articles

Found 2 Documents
Search

Modeling Prevalence of Hypertension in Indonesia with Multivariate Adaptive Regression Splines Method Suliyanto, Suliyanto; Saifudin, Toha; Naura, Sheila Sevira Asteriska; Dewanty, Sanda Insania; Wulandari, Indana Zulfa; Aflaha, Nabila Shafa; Aulia, Niswa Faizah
JTAM (Jurnal Teori dan Aplikasi Matematika) Vol 9, No 2 (2025): April
Publisher : Universitas Muhammadiyah Mataram

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.31764/jtam.v9i2.28392

Abstract

Hypertension is one of the important public health problems in Indonesia, which contributes to the high prevalence of non-communicable diseases. This study aims to model the prevalence of hypertension in Indonesia using the Multivariate Adaptive Regression Splines (MARS) method to identify significant predictors and their interactions. The data used was secondary data from the 2023 Indonesian Health Survey, including variables such as smoking prevalence, physical inactivity, dietary habits (consumption of fatty and sweet foods), lack of fruit and vegetable consumption, and obesity prevalence. The MARS method was used to analyse the nonlinear relationships and interactions between these predictors. After a trial-and-error process to determine the optimal number of basis functions (BF), maximum interactions (MI), and minimum observations (MO), the best model was achieved with BF = 18, MI = 3, and MO = 1. This model produced a Generalised Cross Validation (GCV) value of 13.428 and R-Square of 0.278. This fairly low R-Square value indicates that the factors analysed have contributed to the variation in hypertension prevalence, but there are still other aspects that can be taken into account to improve the predictive power of the model. The significant predictor variables were consumption of fatty foods (X3), lack of physical activity (X2), and consumption of sweets (X4), with the highest importance on X3 (100%). The findings reveal that interactions between variables, such as dietary habits and physical inactivity, significantly influence the prevalence of hypertension. For example, higher consumption of fatty and sweet foods combined with low physical activity increases the risk of hypertension. These results demonstrate the effectiveness of the MARS method in capturing complex and nonlinear relationships and serve as findings that highlight the need for health policies that focus on healthy diets and increased physical activity, in line with Goal 3 of the SDGs, “Good Health and Well-Being,” which aims to reduce premature mortality from noncommunicable diseases. Recommended interventions include nutrition education campaigns and community-based exercise programs to reduce the prevalence of hypertension in Indonesia.
Comparing Multivariate Adaptive Regression Splines and Machine Learning Methods for Classifying Pneumonia in Indonesian Toddlers Kurniawan, Ardi; Azizah, Nur; Naura, Sheila Sevira Asteriska
CAUCHY: Jurnal Matematika Murni dan Aplikasi Vol 10, No 1 (2025): CAUCHY: JURNAL MATEMATIKA MURNI DAN APLIKASI
Publisher : Mathematics Department, Universitas Islam Negeri Maulana Malik Ibrahim Malang

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.18860/cauchy.v10i1.32418

Abstract

Pneumonia is a type of infectious and contagious respiratory disease that causes death in toddlers. According to the Indonesian Ministry of Health (2024), the coverage of pneumonia among toddlers in 2023 was 36.95% with a total of 416,435 cases. This study aims to model and classify the pneumonia status of toddlers in Indonesia using the Multivariate Adaptive Regression Splines (MARS) method and several machine learning methods, such as logistic regression, K-NN, random forest, and SVM. This study uses secondary data from the Survei Kesehatan Indonesia in 2023 and the Profil Kesehatan Indonesia in 2023, including variables such as the percentage of toddlers health service coverage, low birth weight babies, population density, percentage of malnutrition in toddlers, prevalence of smoking in the population aged ≥10 years in the last 1 month, percentage of toddlers who are exclusively breastfed, and percentage of toddlers who have incomplete basic immunization. The best model obtained using the MARS method is with BF = 14, MI = 2, and MO = 3. This model produces a GCV value of 0.122 and R-Square of 82.9%, which shows good prediction performance. The classification results show that the MARS method is superior to the logistic regression, K-NN, random forest, and SVM methods with an accuracy rate of 97.06%.