Article Per Year (5 Year)
p-Index From 2021 - 2026
0.702
P-Index
This Author published in this journals
All Journal JSR : Jaringan Sistem Informasi Robotik Jurnal Ilmu Komputer dan Sistem Komputer Terapan (JIKSTRA) Journal of Artificial Intelligence and Engineering Applications (JAIEA) Journal of Software Engineering and Information System (SEIS)
Marta, Puji Pramudya
Unknown Affiliation
Automotive Engineering Computer Science & IT Control & Systems Engineering Decision Sciences, Operations Research & Management Engineering

Published : 4 Documents Claim Missing Document
Claim Missing Document
Check
Articles

Found 4 Documents
Search

 1 
MODEL PREDIKSI PENERIMA BANTUAN SOSIAL BERBASIS ALGORITMA RANDOM FOREST Sukma, Siti Hatmara; Suarna, Nana; Bahtiar, Agus; Marta, Puji Pramudya; Anam, Khaerul
Jurnal Rekayasa Perangkat Lunak dan Sistem Informasi Vol. 6 No. 1 (2026)
Publisher : Department of Information System Muhammadiyah University of Riau

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.37859/seis.v6i1.10526

Abstract

Inaccurate targeting of social assistance beneficiaries remains a critical issue at the village level due to subjective and inconsistent manual verification processes. This study aims to develop a predictive model for determining social assistance eligibility using the Random Forest algorithm based on 2021 SDGs Village microdata from Cibeureum Village. The research involves data preprocessing, model training, and hyperparameter optimization, with performance evaluation using accuracy, precision, recall, and F1-score metrics. The proposed model achieved an accuracy of 94.34%, indicating strong and stable classification performance. Feature importance analysis shows that housing conditions, access to clean water, and asset ownership are the most influential socioeconomic indicators. These findings demonstrate that Random Forest can effectively support data-driven decision-making and improve the accuracy of social assistance distribution at the village level.
Application of Weighted Loss Function in Convolutional Neural Network for Acne Image Classification Abubakar Sidik; Purnamasari, Ade Irma; Pratama, Denni; Marta, Puji Pramudya; Wijaya, Yudhistira Arie
Journal of Artificial Intelligence and Engineering Applications (JAIEA) Vol. 5 No. 2 (2026): February 2026
Publisher : Yayasan Kita Menulis

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.59934/jaiea.v5i2.1885

Abstract

Automated acne image classification using Convolutional Neural Networks (CNN) holds significant potential in dermatological diagnosis but faces a fundamental challenge of class imbalance. This phenomenon causes standard models to be biased towards majority classes and fail to recognize clinically important minority classes. This study aims to address this bias by applying a Weighted Loss Function to the EfficientNetB1 architecture. The research method employs a comparative experimental approach between two scenarios: the Baseline model (Standard Cross-Entropy) and the Proposed model (Weighted Cross-Entropy). The dataset consists of 5 acne classes with an imbalanced distribution. The results show that the Weighted Loss model significantly outperforms the Baseline model. Overall accuracy increased from 80% to 86%. The most significant improvement occurred in the minority class 'Papules', where the F1-Score surged by 0.10 points (from 0.71 to 0.81). It is concluded that the application of Weighted Loss Function effectively overcomes bias due to imbalanced data without the need for synthetic data augmentation, resulting in a fairer and more reliable model for clinical implementation.
Algoritma LightGBM dengan SMOTE & ADASYN untuk Prediksi Risiko Serangan Jantung Sugianto, Nanda Putri; Purnamasari, Ade Irma; Pratama, Denni; Marta, Puji Pramudya; Wijaya, Yudhistira Arie
JSR : Jaringan Sistem Informasi Robotik Vol 10, No 1 (2026): JSR : Jaringan Sistem Informasi Robotik
Publisher : AMIK Mitra Gama

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.58486/jsr.v10i1.633

Abstract

Ketidakseimbangan data merupakan tantangan utama dalam pemodelan prediksi medis, termasuk prediksi serangan jantung, karena jumlah kasus positif jauh lebih sedikit dibandingkan kasus negatif sehingga menurunkan kemampuan model dalam mendeteksi pasien berisiko tinggi. Penelitian ini bertujuan untuk membandingkan efektivitas dua teknik oversampling, yaitu Synthetic Minority Oversampling Technique (SMOTE) dan Adaptive Synthetic Sampling (ADASYN), dalam meningkatkan performa algoritma Light Gradient Boosting Machine (LightGBM) untuk prediksi risiko serangan jantung. Dataset berjumlah 1.319 sampel dengan sembilan fitur klinis dan dianalisis melalui tahapan pra-pemrosesan, normalisasi, penanganan class imbalance, pembangunan model, serta evaluasi menggunakan Accuracy, Precision, Recall, F1-Score, dan AUC-ROC. Hasil menunjukkan bahwa model baseline memiliki akurasi tinggi namun sensitivitas terhadap kelas positif masih rendah. Setelah diterapkan oversampling, model mengalami peningkatan signifikan. LightGBM-SMOTE memperoleh F1-Score terbesar (0.9876) dan AUC-ROC 0.9853, sedangkan LightGBM-ADASYN mencapai F1-Score 0.9855 dan AUC-ROC 0.9861. Temuan ini menunjukkan bahwa SMOTE memberikan peningkatan performa yang lebih stabil dalam mendeteksi kelas minoritas. Dengan demikian, teknik oversampling khususnya SMOTE terbukti efektif untuk meningkatkan akurasi dan sensitivitas model prediksi serangan jantung.
Klasifikasi Sentimen Ulasan Ipusnas Menggunakan Metode Natural Language Processing dan Support Vector Machine Mustafa, Akmal; Kurniawan, Rudi; Nurhakim, Bani; Marta, Puji Pramudya; Anam, Khaerul
Jurnal Ilmu Komputer dan Sistem Komputer Terapan (JIKSTRA) Vol. 8 No. 1 (2026): Edisi April
Publisher : Universitas Harapan Medan

Show Abstract | Download Original | Original Source | Check in Google Scholar

Abstract

This study investigates sentiment classification of user reviews on the iPusnas digital library application to provide an objective overview of service quality and user experience. Numerous complaints related to login failures, application crashes, and issues in accessing digital books indicate the need for a computational approach capable of processing large volumes of user feedback. The proposed method integrates Natural Language Processing (NLP) techniques with the Support Vector Machine (SVM) algorithm. The workflow consists of collecting 2,000 reviews, applying text cleaning and normalization, tokenization, stopword removal, stemming, rating-based sentiment annotation, and feature extraction using TF-IDF. The dataset was divided using a train–test split for model training and evaluation. Experimental results show that the SVM model achieves 90.1% accuracy, demonstrating strong performance in detecting negative sentiments and moderate performance for positive sentiments due to class imbalance. These  findings highlight the effectiveness of NLP and SVM for extracting user perceptions and indicate the potential of this model as a decision-support tool for improving iPusnas application services. Overall, the study contributes to the advancement of digital service innovation in Indonesia.
Co-Authors Abubakar Sidik Khaerul Anam Mustafa, Akmal Nurhakim, Bani Pratama, Denni Purnamasari, Ade Irma Rudi Kurniawan Suarna, Nana Sugianto, Nanda Putri Sukma, Siti Hatmara Yudhistira Arie Wijaya