Prediksi Risiko Stunting pada Anak Menggunakan Metode Group Method of Data Handling - Type Neural Network

Alif, M. Iqbal (2024) Prediksi Risiko Stunting pada Anak Menggunakan Metode Group Method of Data Handling - Type Neural Network. Other thesis, Institut Teknologi Sepuluh Nopember.

Text 05211740007003-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only until 1 April 2026. Download (2MB) | Request a copy

Abstract

Stunting dianggap sebagai masalah utama kesehatan masyarakat di Indonesia seperti halnya di negara berkembang lainnya. Tingkat prevalensi anak stunting di Indonesia pada tahun 2019 adalah 27,7%. Angka ini masih jauh dari nilai standar prevalensi anak stunting dari WHO yaitu di bawah 20%. Anak stunting dapat menderita kerusakan dan kognitif parah dan tidak dapat diperbaiki, yang menyertai pertumbuhan sang anak terhambat. Efek kerusakan dari stunting dapat bertahan seumur hidup dan bahkan memengaruhi generasi berikutnya. Stunting selama masa kanak-kanak dapat mengakibatkan efek kesehatan negatif di sepanjang usia, termasuk morbiditas dan mortalitas yang tinggi, seperti komplikasi yang mengancam jiwa selama persalinan, peningkatan angka kematian anak, penurunan kinerja dan perkembangan kognitif, peningkatan risiko infeksi, kemiskinan, perkembangan psikomotorik yang tertunda, prestasi sekolah yang lebih rendah, kecerdasan intelektual (IQ) yang rendah, munculnya penyakit kronis, berkurangnya kapasitas produksi di masa dewasa, dengan kerugian dalam pertumbuhan ekonomi dan perkembangan sosial suatu negara. Dalam tugas akhir ini, prediksi risiko terjadinya stunting pada anak di Indonesia dibangun dan diimplementasikan menggunakan model prediksi berbasis jaringan saraf tiruan dengan algoritma Diverse Classifiers Ensemble Based on Group Method of Data Handling (dce-GMDH). dce-GMDH merupakan algoritma GMDH yang menggabungkan berbagai metode pengklasifikasi yang terkenal, seperti support vector machines, random forest, naive Bayes, elastic net logistic regression, dan artificial neural network. Prediksi risiko anak terkena stunting didapatkan berdasarkan kelas status stunting sebagai variabel dependen dan beberapa variabel lainnya, seperti jenis kelamin, tinggi badan anak, berat badan anak, berat badan lahir, usia anak, status imunisasi, tinggi ibu, banyak anak, usia ketika melahirkan, usia kehamilan, pendidikan ibu, konsumsi tablet zat besi, tempat tinggal, dan kondisi lingkungan sebagai variabel independen. Nilai variabel yang diperlukan untuk pemodelan dce-GMDH diperoleh dari data yang berasal dari Indonesian Family Life Survey (IFLS). Hasil eksperimen terhadap model terbaik yang didapatkan adalah model dce-GMDH dengan parameter kriteria eksternal MSE (Mean Square Error) dan melibatkan beberapa variabel independen yang telah melalui proses seleksi variabel (jenis kelamin, usia anak, tinggi badan anak, berat badan anak, berat badan lahir anak, tinggi badan ibu, usia ibu saat melahirkan, dan tempat tinggal). Performa yang diperoleh dari model terbaik tersebut adalah akurasi sebesar 88,71%, sensitivitas sebesar 92,52%, dan spesifisitas sebesar 85,28%.
=================================================================================================================================
Stunting is considered a major public health problem in Indonesia as in other developing countries. The prevalence rate of stunted children in Indonesia in 2019 was 27.7%. This prevalence rate is still far from the WHO standard value for the prevalence of stunted children, which is below 20%. Stunted children can suffer severe and irreversible cognitive damage that accompanies the child's stunted growth. The damaging effects of stunting can last for a lifetime and even affect the next generation. Stunting during childhood can result in negative health effects throughout the lifespan, including high morbidity and mortality, such as life-threatening complications during childbirth, increased child mortality, reduced cognitive performance and development, increased risk of infection, poverty, delayed psychomotor development, lower school performance, lower intellectual quotient (IQ), emergence of chronic diseases, reduced productive capacity in adulthood, with losses in a country's economic growth and social development. In this final project, prediction of the risk of stunting in children in Indonesia is built and implemented using an artificial neural network-based prediction model with the Diverse Classifiers Ensemble Based on Group Method of Data Handling (dce-GMDH) algorithm. dce-GMDH is a GMDH algorithm that combines various well-known classifier methods, such as support vector machines, random forests, naive Bayes, elastic net logistic regression, and artificial neural networks. Predictions of a child's risk of stunting are obtained based on stunting status class as the dependent variable and several other variables, such as gender, child height, child weight, birth weight, child age, immunization status, mother's height, number of children, age at birth, length of pregnancy, maternal education, consumption of iron tablets, place of residence, and environmental conditions as independent variables. The variable values required for dce-GMDH modeling were obtained from data originating from the Indonesian Family Life Survey (IFLS). The experimental results of the best model obtained were the dce-GMDH model with the external criterion parameter MSE (Mean Square Error) and involving several independent variables that had gone through a variable selection process (gender, child's age, child's height, child's weight, child’s birth weight, mother's height, mother's age at birth, and place of residence). The performance obtained from the best model was accuracy of 88,71%, sensitivity of 92,52%, and specificity of 85,28%.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	prediksi, stunting, jaringan saraf tiruan, GMDH, prediction, stunting, artificial neural network
Subjects:	R Medicine > RJ Pediatrics > RJ101 Child Health. Child health services T Technology > T Technology (General) > T57.84 Heuristic algorithms.
Divisions:	Faculty of Intelligent Electrical and Informatics Technology (ELECTICS) > Information System > 57201-(S1) Undergraduate Thesis
Depositing User:	M. Iqbal Alif
Date Deposited:	12 Feb 2024 08:56
Last Modified:	12 Feb 2024 08:56
URI:	http://repository.its.ac.id/id/eprint/106920

Actions (login required)

View Item