Claim Missing Document
Check
Articles

Found 3 Documents
Search
Journal : JUSTIN (Jurnal Sistem dan Teknologi Informasi)

Early Detection of Stunting in Toddlers Based on Ensemble Machine Learning in Purbaratu Tasikmalaya AL Husaini; Irani Hoeronis; Hen Hen Lumana; Luh Desi Puspareni
JUSTIN (Jurnal Sistem dan Teknologi Informasi) Vol 11, No 3 (2023)
Publisher : Jurusan Informatika Universitas Tanjungpura

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.26418/justin.v11i3.66465

Abstract

This research utilizes combines several algorithm model that improve the accuracy of early detection of stunting in toddlers in Purbaratu Tasikmalaya.  The ensemble method used a voting classifier to combine the prediction results of models. The data used in this research were anthropometric data from 195 toddlers in Purbaratu Tasikmalaya. Results of the testing have identified that the use of the ensemble model machine learning method produces high accuracy for 3 categories of anthropometric data categories tested, that combined accuracy value 97,43 %, 92,30%, and 94,87% for all ensemble model and category.
Rancang Bangun Sistem Informasi Deteksi Dini Stunting dengan Metode Artificial Neural Network Lukmana, Hen Hen; Al-Husaini, Muhammad; Puspareni, Luh Desi; Hoeronis, Irani
JUSTIN (Jurnal Sistem dan Teknologi Informasi) Vol 12, No 3 (2024)
Publisher : Jurusan Informatika Universitas Tanjungpura

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.26418/justin.v12i3.80119

Abstract

Stunting pada anak merupakan masalah kesehatan malnutrisi kronis yang menjadi perhatian serius di Indonesia.  Stunting dapat terjadi pada anak yang mengalami kekurangan gizi kronis, terutama pada usia 0-23 bulan. Faktor-faktor yang menyebabkan stunting pada anak sangat kompleks dan melibatkan berbagai faktor seperti gizi, kesehatan, sosial ekonomi, lingkungan, genetik dan peilaku. Penelitian ini bertujuan untuk merancang dan mengembangkan sistem informasi deteksi dini stunting menggunakan teknologi artificial neural network yang dilengkapi dengan stacking classifiers dengan dikombinasikan ensemble machine learning gradient boosting, random forest dan output estimator regresi logistik, selain itu pengembangan sistem ini dilakukan dengan menggunakan metode pengembangan waterfall. Sistem ini diharapkan dapat memprediksi risiko stunting secara akurat berdasarkan data pertumbuhan anak, serta memberikan rekomendasi intervensi yang tepat. Penggunaan neural network memungkinkan analisis data yang kompleks dan pembaruan model secara berkala dengan hasil rataan akurasi prediksi kombinasi beberapa algoritma menggunakan model stacking classifiers dan cross validation tersebut menghasilkan akurasi yang stabil di 86,22% berdasarkan dataset 10 ribu label target prediksi. Hasil dari penelitian berdasarkan model pengembangan dan pelatihan model ini mencakup analisis kebutuhan sistem, perancangan dan desain sistem dengan UML, implementasi sistem dengan fitur pengecekan stunting, artikel edukasi, konsultasi, login dan registrasi, dan hasil pengujian dengan System Usability Scale (SUS) dengan nilai rata-rata 81 yang termasuk pada grade A dan blackbox testing dengan hasil sesuai harapan.
Ulcerative Colitis Classification on Endoscopy Image using Support Vector Machine with Image Extraction using Gray Level Co-Occurrence Matrix Nurrohman, Agni; Hoeronis, Irani; Lukmana, Hen Hen
JUSTIN (Jurnal Sistem dan Teknologi Informasi) Vol 12, No 4 (2024)
Publisher : Jurusan Informatika Universitas Tanjungpura

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.26418/justin.v12i4.82903

Abstract

Ulcerative colitis or inflammation of the colon is a chronic inflammatory disorder characterized by mucosal inflammation involving the large intestine (colon) and leading to the anus (rectum). The number of cases of ulcerative colitis ranges from 90-505 people out of 100,000 people in Northern Europe and North America, less common in Western and Southern European regions as well as at least 10 times less in Asia, Africa and Oriental populations. This study aims to classify endoscopic images with the Support Vector Machine method with the results of feature extraction using Gray Level Co-Occurrence Matrix. The dataset used is the kvasir dataset with the number of datasets used in this study totaling 1990 with each class, namely the healthy class and the ulcerative colitis class, having 995 images. Endoscopy results in the form of digital images captured using a small camera inserted into the patient's gastrointestinal tract. In this study, the accuracy model of Ulcerative Colitis classification was calculated using the results of endoscopy image feature extraction with GLCM feature extraction using SVM classification with RBF kernel. The search for hyperparameter values is carried out to find the best C and gamma values so that this study has model accuracy results which previously had an accuracy of 86.45% to 90.85%, a precision value of 91.58%, a recall value of 90.68% and an f1-score value of 91.12%.