Article Per Year (5 Year)
p-Index From 2021 - 2026
2.195
P-Index
This Author published in this journals
All Journal Jurnal Pendidikan Teknologi dan Kejuruan KARMAPATI (Kumpulan Artikel Mahasiswa Pendidikan Teknik Informatika) ISSN: 2252-9063 Journal of ICT Research and Applications Jurnal Informatika dan Teknik Elektro Terapan Knowledge Engineering and Data Science JAPANEDU: Jurnal Pendidikan dan Pengajaran Bahasa Jepang Jurnal Nasional Pendidikan Teknik Informatika (JANAPATI) JURIKOM (Jurnal Riset Komputer) Jurnal Informatika dan Rekayasa Elektronik Journal of Computer System and Informatics (JoSYC) JIKA (Jurnal Informatika) Jurnal Pengabdian UNDIKMA International Journal of Management Science and Information Technology (IJMSIT) Jurnal Informatika Polinema (JIP) Jurnal Nasional Teknologi Informasi dan Aplikasinya
Putu Hendra Suputra
Universitas Pendidikan Ganesha
Religion Agriculture, Biological Sciences & Forestry Arts Humanities Computer Science & IT Control & Systems Engineering Decision Sciences, Operations Research & Management Economics, Econometrics & Finance Education Electrical & Electronics Engineering Engineering Environmental Science Industrial & Manufacturing Engineering Languange, Linguistic, Communication & Media Law, Crime, Criminology & Criminal Justice Library & Information Science Mathematics Nursing Public Health Social Sciences Other

Published : 23 Documents Claim Missing Document
Claim Missing Document
Check
Articles

Found 23 Documents
Search

 1   2   3 
KLASIFIKASI PENYAKIT DAUN TOMAT BERBASIS TRANSFER LEARNING EFFICIENTNETB0 Sapanca, Ganeis; Suputra, Putu Hendra; Ni Wayan, Marti
Jurnal Informatika dan Rekayasa Elektronik Vol. 9 No. 1 (2026): JIRE April 2026
Publisher : LPPM STMIK Lombok

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.36595/jire.v9i1.2006

Abstract

Penyakit daun tomat seperti Early Blight dan Late Blight menjadi ancaman serius terhadap produktivitas pertanian yang memerlukan deteksi dini untuk mencegah kerugian hasil panen. Penelitian ini mengembangkan sistem klasifikasi penyakit daun tomat menggunakan arsitektur EfficientNetB0 dengan metode transfer learning untuk mengidentifikasi tiga kategori kondisi daun Early Blight, Healthy, dan Late Blight. Dataset terdiri dari 1.425 gambar yang dibagi dengan rasio 80:20 dan ditingkatkan menjadi 4.557 gambar training melalui teknik augmentasi data. Empat skenario eksperimen dilakukan dengan variasi batch size (16 dan 32) serta jumlah epoch (25 dan 50) untuk mengidentifikasi konfigurasi hyperparameter optimal. Hasil evaluasi menunjukkan bahwa model dengan batch size 16 dan epoch 50 menghasilkan performa terbaik dengan test accuracy 92.55%, precision 92.82%, recall 92.55%, dan F1-score 92.39%. Performa per-kelas mencapai Early Blight (F1-score 92.46%), Healthy (F1-score 96.26% dengan recall sempurna 100%), dan Late Blight (F1-score 88.30%). Analisis komparatif membuktikan batch size 16 konsisten unggul dibanding batch size 32 dengan selisih akurasi 4.21%, dan peningkatan epoch dari 25 ke 50 memberikan kenaikan 2.10% terutama pada deteksi Early Blight. Model yang dikembangkan memenuhi standar aplikasi praktis dengan akurasi tinggi dan waktu training efisien, berpotensi membantu petani melakukan deteksi dini penyakit untuk meningkatkan produktivitas tanaman tomat.
Continuous Regression Models for Mapping the Smartphone Addiction Spectrum Using Random Forest Regressor Eka Aditya Saputra; I Made Gede Sunarya; Putu Hendra Suputra
International Journal of Management Science and Information Technology Vol. 6 No. 1 (2026): January - June 2026
Publisher : Lembaga Komunitas Informasi Teknologi Aceh (KITA), Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.35870/ijmsit.v6i1.7098

Abstract

This study proposes a predictive modeling approach to measure the level of smartphone addiction in adolescents by transforming a conventional binary classification model into a continuous regression model. The use of categorical labels often fails to capture the complex spectrum of addictive behaviors, so this study implemented the Random Forest Regressor algorithm to predict addiction scores on a scale of 1.0 to 10.0. The experimental results show that the regression model is able to provide high prediction accuracy, as evidenced by the coefficient of determination obtained R^2 of 0.8607 and a Mean Absolute Error (MAE) of 0.2854. These findings confirm that the regression approach offers better data resolution in mapping the degree of digital dependency than classification methods. In practice, this model produces a continuous score that provides a dynamic tool for mental health professionals. This approach allows for objective monitoring of patient’s behavioral progress during recovery. Furthermore, this model can facilitate multilevel psychological interventions and tailored care, from early prevention to therapy for high-risk addicts.
Comparative Analysis of Machine Learning Models for Early Dengue Detection Using Clinical Symptoms Made Dwi Aprillia Kusuma Wiryani; I Gusti Ayu Agung Diatri Indradewi; Putu Hendra Suputra
KARMAPATI (Kumpulan Artikel Mahasiswa Pendidikan Teknik Informatika) Vol. 15 No. 2 (2026): [ONGOING] Karmapati Vol 15 No 2 Tahun 2026
Publisher : Universitas Pendidikan Ganesha

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.23887/karmapati.v15i2.115451

Abstract

Early detection of dengue fever remains a clinical challenge without prior laboratory examination, due to its overlapping symptoms with other febrile conditions, most notably high body temperature. By observing these symptom similarities, this study proposes a comprehensive comparison of machine learning algorithms using a classification-based early detection approach with Random Forest, Gradient Boosting, Support Vector Machine, and Decision Tree. The dataset was obtained from a regional hospital with ethical clearance, consisting of 212 records collected in January 2025, with 11 selected clinical features including body temperature, fever duration, pain, nausea, vomiting, cough, flu, rash, headache, nosebleed, and gum bleed. The models were evaluated using accuracy, precision, recall, and F1-score, both before and after hyperparameter tuning. GridSearchCV was applied to identify the optimal hyperparameter combination for each model. In this study, recall is prioritized as the primary evaluation metric to ensure the model performs well in minimizing missed dengue cases. The results demonstrated that SVM achieved the best performance across all metrics except precision, both before and after tuning. These findings suggest that SVM is the most suitable model for clinical early detection of dengue fever using symptom-based data.
Co-Authors Ardyaputra, Gede Yudha Aryadi, Made Waradiana Desak Made Sri Mardani Eka Aditya Saputra Eko Mulyanto Yuniarno Gijsbertus Jacob Verkerke I Gusti Ayu Agung Diatri Indradewi I Gusti Putu Kresna Dana I Kadek Antartika I Ketut Eddy Purnama I Ketut Purnamawan I Ketut Purnamawan I Ketut Resika Arthana I Made Ardwi Pradnyana I Made Candiasa I Made Gede Sunarya I Md. Dendi Maysanjaya I Wayan Sadyana I Wayan Sadyana Ida Bagus Nyoman Pascima Komang Setemen Laksmi, Ida Ayu Luh Joni Erawati Dewi Made Dwi Aprillia Kusuma Wiryani Made Windu Antara Kesiman Myrtati Dyah Artaria Ni Ketut Kertiasih Ni Wayan Marti Ni Wayan, Marti Pratama Putra, Gede Bakti Putra, Gede Bakti Pratama Sapanca, Ganeis Sensusiati, Anggraini Dwi Sugiantari, Kadek Feny