Article Per Year (5 Year)
p-Index From 2021 - 2026
1.829
P-Index
This Author published in this journals
All Journal Dinamik Jurnal Riset Informatika JOURNAL OF INFORMATION SYSTEM RESEARCH (JOSH) Jurnal Sistem Komputer dan Informatika (JSON) Journal of Computer Science, Information Technology and Telecommunication Engineering (JCoSITTE) Journal of Computer Networks, Architecture and High Performance Computing International Journal of Community Service KLIK: Kajian Ilmiah Informatika dan Komputer Jurnal Sains Teknologi dan Sistem Informasi Data Sciences Indonesia (DSI)
Agung Prabowo
Faculty Of Science And Technology, Universitas Prima Indonesia, Medan, Indonesia
Religion Humanities Computer Science & IT Control & Systems Engineering Decision Sciences, Operations Research & Management Education Electrical & Electronics Engineering Environmental Science Languange, Linguistic, Communication & Media Law, Crime, Criminology & Criminal Justice Library & Information Science Physics Public Health Social Sciences Other

Published : 10 Documents Claim Missing Document
Claim Missing Document
Check
Articles

Found 10 Documents
Search

 1 
Komparasi Tingkat Akurasi Random Forest dan Decision Tree C4.5 Pada Klasifikasi Data Penyakit Infertilitas Agung Prabowo; Sumita Wardani; Rico Wijaya Dewantoro; Wilfredo Wesly; Leonardo
KLIK: Kajian Ilmiah Informatika dan Komputer Vol. 4 No. 1 (2023): Agustus 2023
Publisher : STMIK Budi Darma

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.30865/klik.v4i1.1115

Abstract

Male fertility has declined over the past two decades. The decrease is due to environmental factors, such as lifestyle habits that can affect the quality of a man's sperm. Artificial intelligence technology is currently developing as a methodology for health decision support systems. In the process of predicting infertility can be done by applying Machine Learning technology. This study focuses on comparing the Random Forest classification method with Decision Tree C4.5 to see the level of accuracy in predicting the success of infertility data classification. Data for the Fertility Dataset was obtained from the UCI Machine Learning Repository with a total of 100 data records, 10 attributes and 2 attribute classes, namely Normal and Altered. The parameters used are age, childhood diseases, accidents or trauma, surgical operations, alcohol consumption and smoking habits. Then evaluate the testing of the two methods, namely by using 10fold Cross Validation. Based on the results of Random Forest and Decision Tree C4.5 testing, the average accuracy of Random Forest is 87.20% and Decision Tree C4.5 with an accuracy rate of 85.90%. From the results obtained, it can be concluded that Random Forest is a superior method by 1.3% when compared to Decision Tree C4.5 in predicting accuracy in the Fertility Dataset.
Diagnosis and Prediction of Chronic Kidney Disease Using a Stacked Generalization Approach Agung Prabowo; Sumita Wardani; Abdul Muis; Radiman Gea; Nathanael Atan Baskita Tarigan
Journal of Computer Networks, Architecture and High Performance Computing Vol. 6 No. 1 (2024): Article Research Volume 6 Issue 1, January 2024
Publisher : Information Technology and Science (ITScience)

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47709/cnahpc.v6i1.3611

Abstract

Chronic Kidney Disease (CKD) is. In the past, several learners have been applied for prediction of CKD but there is still enough space to develop classi?ers with higher accuracy. The study utilizes chronic kidney disease dataset from UCI Machine Learning Repository. In this paper, individual approaches, viz., linear-SVM, kernel methods including polynomial, radial basis function, and sigmoid have been used while among ensembles majority voting and stacking strategies have been applied. Stacked Ensemble is based on various types of meta-learners such as C4.5, NB, k-NN, SMO, and logit-boost. The stacking approach with meta-learner Logit-Boost (ST-LB) achieves accuracy 98,50%, sensitivity 98,50%, false positive rate 20,00%, precision 98,50%, and F-measure 98,50% demonstrating that it is the best classi?er as compared to any of the individual and ensemble approaches
Analisis Metode WASPAS dalam Menentukan Pengangkatan Pegawai Kontrak Menjadi Pegawai Tetap Abdul Muis; Abdul Muis; Akbar Idaman; Handry Eldo; Agung Prabowo; Ryan Rinaldi Hadistio
SATESI: Jurnal Sains Teknologi dan Sistem Informasi Vol. 4 No. 1 (2024): April 2024
Publisher : Yayasan Pendidikan Penelitian Pengabdian ALGERO

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.54259/satesi.v4i1.3034

Abstract

Human Resources (HR) is a valuable asset for every company, and good management greatly affects operational success. PT XYZ faces challenges in the process of appointing contract employees to permanent employees which is currently done manually, causing a slow and less accurate process. This research aims to develop a method that simplifies and accelerates the process using Weight Aggregated Sum Product Assessment (WASPAS) to simplify and speed up the decision-making process in appointing contract employees to become permanent employees at PT. XYZ, so that decisions taken can be faster, more precise and accurate. This method was chosen for its ability to reduce errors and optimize the assessment with various criteria. The results showed that Alternative 9 was ranked first with a Qi value of 0.9685, showing the best performance among other candidates. The implementation of the WASPAS method is expected to help PT XYZ in making faster, more precise, and accurate decisions, thereby increasing efficiency and objectivity in employee hiring, as well as improving the performance and stability of the company's human resources.
Implementation of Complex Propotional Assesment (COPRAS) in Determining Air Conditioning System Traders Idaman, Akbar; Amrullah, Amrullah; Raja Gunung, Tar Muhammad; Eldo, Handry; Prabowo, Agung
Journal of Computer Science, Information Technology and Telecommunication Engineering Vol 5, No 2 (2024)
Publisher : Universitas Muhammadiyah Sumatera Utara, Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.30596/jcositte.v5i2.19417

Abstract

Increased global warming and awareness of the need to reduce greenhouse gas emissions have strengthened the focus on energy efficiency in various sectors, including the HVAC (Heating, Ventilation, and Air Conditioning) industry. In this context, the selection of air conditioning (AC) systems becomes crucial in providing thermal comfort. However, decisions regarding air conditioning systems are often complex as they involve considerations of energy efficiency, operational costs, system reliability, and environmental impact. To address this complexity, Complex Proportional Assessment (COPRAS) emerged as an effective multi-criteria analysis method. However, the application of COPRAS in determining AC system traffickers is still limited. This research explores the possible application of COPRAS in this context and identifies key factors to consider. The evaluation results show that Medan Elektronik and Citra Inovasi Prima are the top choices in the selection of AC system traffickers. This research is expected to contribute to the development of more sophisticated analysis methods in the HVAC industry as well as assist decision makers in selecting more appropriate and sustainable air conditioning systems.
Perbandingan Kinerja Algoritma Random Florest Classifier Dan Lightgbm Classifier Untuk Prediksi Penyakit Jantung Duran, Filbert; Wijaya, Frederico; Hulu, Yakin Rianto; Harahap, Mawaddah; Prabowo, Agung
Data Sciences Indonesia (DSI) Vol. 3 No. 2 (2023): Article Research Volume 3 Issue 2, December 2023
Publisher : Yayasan Cita Cendikiawan Al Kharizmi

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47709/dsi.v3i2.3831

Abstract

Penyakit jantung merupakan masalah kesehatan serius yang dapat dicegah dan diobati. Dengan menjaga gaya hidup sehat, melakukan pemeriksaan kesehatan secara rutin, dan mengikuti anjuran dokter[1], risiko penyakit jantung dapat dikurangi. Random Forest Classifier (RFC) bagaikan hutan pohon keputusan yang bekerja sama untuk menghasilkan prediksi yang lebih jitu. Algoritma ini tergolong handal dan fleksibel, mampu menangani berbagai tugas klasifikasi dan regresi. Kelebihannya, RFC menawarkan akurasi tinggi, tahan terhadap overfitting, dan mudah diinterpretasikan[2]. RFC adalah algoritma machine learning yang kuat dengan banyak keunggulan, namun perlu dipertimbangkan pula keterbatasannya dalam hal komputasi dan fleksibilitas[3]. LightGBM merupakan algoritma machine learning yang kuat dan efisien untuk klasifikasi dan regresi. Kecepatan, akurasi, dan kemudahan penggunaannya menjadikannya pilihan yang menarik untuk berbagai aplikasi[4]. Dari hasil yang didapat dari penelitian ini adalah metode RFC dan LightGBM dapat disimpulkan bahwa metode RFC merupakan metode yang tergolong efektif dalam analisis penyakit jantung dengan akurasi prediksi dari model adalah 95,37%., dapat dikatakan bahwa metode Random Florest Classifier cocok untuk melakukan analisis penyakit jantung bedasarkan dataset yang ada.
INTEGRATION SYSTEM OF THREE SECURITY FEATURES IN SMART DUAL MCB WITH AUTOMATIC LOAD BALANCING AND FIRE DETECTION BASED ON ARDUINO UNO (SINTAKS) Achmad Ridwan; Agung Prabowo
Jurnal Riset Informatika Vol. 7 No. 3 (2025): Juni 2025
Publisher : Kresnamedia Publisher

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.34288/jri.v7i3.363

Abstract

The increasing use of electronic devices in Indonesian households has significantly strained traditional electrical systems, with electricity consumption growing 4.5% annually and 78% of urban homes utilizing over 10 electronic devices. This situation poses substantial fire risks, as electrical short circuits cause 62.8% of urban fires, with MCB overloads accounting for 27% of incidents. This research introduces SINTAKS (Sistem Integrasi Tiga Keamanan Smart Dual MCB), an innovative integrated system combining three essential safety features: energy monitoring, automatic load balancing, and early fire detection. Unlike conventional systems requiring separate components, SINTAKS provides a comprehensive solution using Arduino Uno as the main controller, integrated with ACS712 current sensors, DS18B20 temperature sensors, MQ-2 smoke detectors, and relay modules. The system demonstrates remarkable performance with 97.8% current measurement accuracy, load balance improvement from 62.7% to 91.3%, and fire detection response time of 2.9-4.7 seconds. Field testing in real household installations confirmed system reliability with 94.8% success rate across various operational scenarios. SINTAKS achieves 4.2% energy savings while maintaining cost-effectiveness at IDR 875,000, making it accessible for widespread residential implementation. This autonomous system operates independently without IoT dependence, ensuring reliable protection even in offline environments. The research successfully addresses critical gaps in household electrical safety through practical, affordable, and integrated technology.
Training and Mentoring on Meta-Product-Based Digital Marketing Strategies to Increase the Competitiveness of School MSMEs Bambang Suwarno; Bernard Ekarisman Ndururu; Feby Yoana Siregar; Okta Jaya Hermaja; Herlin Munthe; Kristi Endah Ndilosa Ginting; Muhammad Fikri Akbar Zuhdi; Puji Syukran; Agung Prabowo
International Journal Of Community Service Vol. 6 No. 2 (2026): May 2026 ( Indonesia - Thailand - Malaysia - Philippines)
Publisher : CV. Inara in Colaboration with www.stie-sampit.ac.id

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.51601/ijcs.v6i2.1007

Abstract

The purpose of this Community Service activity is to improve the understanding and skills of students in grades X, XI, and XII of SMKN 1 Binjai in implementing digital marketing strategies based on Meta products to support the competitiveness of school MSMEs. The implementation method uses a participatory training approach and direct practical mentoring which includes understanding digital marketing concepts, introducing Meta products, creating promotional content, managing customer communications, and evaluating data-based marketing performance. Success was measured through pre-tests, post-tests, group practice observations, and Likert-scale questionnaires to assess the increase in knowledge, skills, and perceptions of participants regarding the benefits of the training. The results of the activity showed an increase in participants' understanding after participating in the training, especially in the aspect of using Instagram, Facebook, WhatsApp Business, and Meta Business Suite as promotional media for school MSMEs. Participants also demonstrated the ability to compose promotional captions, create digital content designs, manage customer communications, and understand performance indicators such as reach, impressions, engagement, clicks, comments, and incoming messages. The results of this training and mentoring have been proven to help students understand digital marketing in a more applicable, creative, and data-driven way. This activity also strengthens students' digital entrepreneurship skills by utilizing the Meta ecosystem as a learning and promotional medium. The limitation of this activity lies in the evaluation, which still focuses on short-term results through pre-tests, post-tests, and participant perceptions. Contributions: This activity provides practical contributions in the form of a digital marketing training model based on Meta products for vocational school students and theoretical contributions in strengthening digital entrepreneurship in schools.
Analisis Sentimen Aplikasi Elektrokardiogram di Play Store Berbasis IndoBERT dan BERTopic Muhammad Chairul Izzat; Daniel Parlindungan Simanjuntak; Fajar Michael Sanjaya Sianturi; Pianus Lase; Agung Prabowo
Jurnal Sistem Komputer dan Informatika (JSON) Vol. 7 No. 4 (2026): Juni 2026
Publisher : Universitas Budi Darma

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.30865/json.v7i4.9767

Abstract

Perkembangan aplikasi kesehatan digital, khususnya aplikasi Elektrokardiogram (EKG), meningkatkan jumlah ulasan pengguna pada Google Play Store yang dapat dimanfaatkan sebagai sumber informasi untuk mengevaluasi kualitas aplikasi. Namun, ulasan pengguna bersifat tidak terstruktur sehingga sulit dianalisis secara manual. Penelitian ini bertujuan menganalisis sentimen dan mengidentifikasi topik utama pada ulasan pengguna aplikasi EKG menggunakan pendekatan Natural Language Processing (NLP). Metode yang digunakan adalah analisis sentimen dengan IndoBERT dan topic modeling dengan BERTopic. Dataset penelitian terdiri dari 1000 ulasan berbahasa Indonesia yang diperoleh melalui proses web scraping. Tahapan penelitian meliputi preprocessing, pembagian data latih dan uji, analisis sentimen, topic modeling, serta penyajian hasil dalam sistem informasi monitoring. Hasil penelitian menunjukkan bahwa model IndoBERT mampu melakukan klasifikasi sentimen dengan akurasi sebesar 72%, yang menunjukkan performa cukup baik, meskipun kemampuan klasifikasi pada sentimen negatif dan netral masih terbatas akibat ketidakseimbangan distribusi data. BERTopic berhasil mengidentifikasi topik utama, meliputi akurasi aplikasi, kemudahan penggunaan, serta kendala teknis seperti error dan masalah kamera. Integrasi kedua metode memberikan analisis yang lebih komprehensif karena mampu menunjukkan kecenderungan sentimen sekaligus isu utama yang dibahas pengguna. Hasil analisis divisualisasikan melalui sistem informasi monitoring sehingga dapat mendukung evaluasi aplikasi berbasis data.
Klasifikasi Tingkat Emisi Karbon Individu menggunakan Support Vector Machine (SVM) berdasarkan Data Perilaku Gaya Hidup Andreas Halawa; Hamdani Hutauruk; Agung Prabowo
Dinamik Vol 31 No 2 (2026)
Publisher : Universitas Stikubank

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.35315/dinamik.v31i2.10571

Abstract

The increase in carbon emissions due to individual activities and lifestyles has become one of the main contributors to global climate change. The emerging issue is how to accurately classify individual carbon impact levels based on lifestyle behavior data. This study aims to apply the Support Vector Machine (SVM) method in classifying carbon emission levels into three categories, namely High, Medium, and Low. The dataset used consists of 1,400 data points with 10 initial features, which include numerical and categorical variables. The preprocessing stage is carried out through standardization using StandardScaler and categorical transformation with OneHotEncoder within a Pipeline framework to prevent data leakage. The SVM model with a Radial Basis Function (RBF) kernel was optimized using GridSearchCV and produced the best parameters C =75 and γ = 0.0588. With an 80:20 data split, the model achieved an accuracy of 94%, and precision, recall, and F1-score values ranged from 90% to 95%. An AUC value of 98%–100% indicates a very good discriminatory ability. The study results conclude that the SVM method is effective and reliable for classifying carbon emission levels based on lifestyle behavior.
Perbandingan Algoritma K-Nearest Neighboor dan Naive Bayes Dalam Prediksi Penyakit Ginjal Kronis Pada Lansia Marco Duran Simbolon; Dimas Dimanta Bukit; Rian Elby Purba; Faisal Haries Ketaren; Agung Prabowo
Journal of Information System Research (JOSH) Vol 6 No 2 (2025): January 2025
Publisher : Forum Kerjasama Pendidikan Tinggi (FKPT)

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47065/josh.v6i2.5938

Abstract

Chronic kidney disease is a serious illness that requires early diagnosis to improve treatment outcomes, especially in the elderly. The main challenge in diagnosing this disease lies in the fact that symptoms often do not appear until the disease has reached an advanced stage, which necessitates the use of accurate prediction methods. Additionally, the dataset's limited size, consisting of only 195 patient records, may affect the algorithm's ability to identify patterns. Choosing the appropriate algorithm is also a challenge, as some algorithms have limitations in handling complex medical data. This study aims to evaluate the performance of the K-Nearest Neighbors (KNN) and Naïve Bayes algorithms in predicting chronic kidney disease. The dataset was analyzed using Weka Waikato software and tested using the 9-fold cross-validation method. The best results were obtained using the Naïve Bayes algorithm, with an accuracy of 97.4359%. Based on these results, it can be concluded that both algorithms can be used to predict chronic kidney disease in the elderly. However, to further improve prediction accuracy, proposed solutions include expanding the dataset with more diverse data and optimizing the algorithm's hyperparameters. On the other hand, the Naïve Bayes algorithm demonstrated higher accuracy compared to KNN in this study, making it the more recommended choice.
Co-Authors ABDUL MUIS Abdul Muis Achmad Ridwan Akbar Idaman Amrullah Andreas Halawa Bambang Suwarno Bernard Ekarisman Ndururu Daniel Parlindungan Simanjuntak Dimas Dimanta Bukit Duran, Filbert Eldo, Handry Faisal Haries Ketaren Fajar Michael Sanjaya Sianturi Feby Yoana Siregar Hamdani Hutauruk Handry Eldo Herlin Munthe Hulu, Yakin Rianto Kristi Endah Ndilosa Ginting Leonardo Marco Duran Simbolon Mawaddah Harahap, Mawaddah Muhammad Chairul Izzat Muhammad Fikri Akbar Zuhdi Nathanael Atan Baskita Tarigan Okta Jaya Hermaja Pianus Lase Puji Syukran Radiman Gea Rian Elby Purba Rico Wijaya Dewantoro Ryan Rinaldi Hadistio Sumita Wardani Tar Muhammad Raja Gunung Wijaya, Frederico Wilfredo Wesly