Claim Missing Document
Check
Articles

Found 19 Documents
Search

Algoritma C4.5 sebagai Penerapan Decision Tree-Based Classification Model untuk Mengklasifikasikan Tingkat Omzet UMKM Berdasarkan Profil Bisnis Yudhi Fajar Saputra; Sugiarto
Jurnal Ekonomika: Manajemen, Akuntansi, dan Perbankan Syari'ah Vol. 13 No. 2 (2024): September
Publisher : Economic Faculty, University of Widya Gama Mahakam Samarinda

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.24903/je.v13i2.3562

Abstract

Micro, Small, and Medium Enterprises (MSMEs) significantly contribute to the national economy. However, many of them experience stagnant revenue due to limited business profiles and the lack of data-driven development strategies. This study aims to build a classification model for MSME revenue levels based on business profile attributes using the Decision Tree C4.5 algorithm. The dataset consists of over 13,000 publicly available records, which were preprocessed and categorized into three revenue classes: low, medium, and high, based on quartile distribution. The results show that the C4.5 model achieves a classification accuracy of 48.53%, with a dominant prediction in the medium revenue category. The resulting decision tree structure generates interpretable and logical rules, such as: “If the business type is services, not legally registered, and has assets less than IDR 7 million, then the revenue tends to be medium.” Further analysis reveals that attributes such as business type, legal status, assets, and production capacity are key predictors of MSME revenue classification. Although the model's accuracy is still limited, this approach provides a solid foundation for developing decision support systems for MSME development agencies. The study recommends exploring additional features and implementing ensemble methods to improve model performance in future research
PENGENALAN ARTIFICIAL INTELLIGENCE KEPADA SISWA DI SMA NEGERI 9 SAMARINDA Aldi Bastiatul Fawait; Sitti Rahmah; Sugiarto Sugiarto; La Jupriadi Fakhri; Muh Jamil; Yudhi Fajar Saputra; Merlina Lidiana Arifin; Nadia Keril Saputri
Community Development Journal : Jurnal Pengabdian Masyarakat Vol. 5 No. 6 (2024): Vol. 5 No. 6 Tahun 2024
Publisher : Universitas Pahlawan Tuanku Tambusai

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.31004/cdj.v5i6.38605

Abstract

Kegiatan pengabdian pada masyarakat yang melibatkan sosialisasi mengenai Pengenalan Artificial Intelligence (AI) kepada siswa di lingkungan Sekolah Menengah Atas ini bertujuan untuk memberikan informasi kepada siswa-siswi SMA Negeri 9 Samarinda mengenai perkembangan teknologi terkait AI. Dalam proses sosialisasi ini, tujuannya adalah memberikan pemahaman mengenai kecerdasan buatan yang dirancang khusus untuk menangani permasalahan kognitif yang umumnya terkait dengan kecerdasan manusia, seperti proses pembelajaran, pemecahan masalah, dan pengenalan pola. Manfaat dari kegiatan pengabdian pada masyarakat ini adalah agar siswa-siswi dapat lebih memahami kemajuan teknologi kecerdasan buatan di Indonesia dan memanfaatkan Sumber Daya Manusia (SDM) lokal yang berkualitas, seiring dengan pemahaman bahwa SDM dalam negeri tidak kalah dengan SDM luar negeri. Selain itu, diharapkan bahwa keberadaan Artificial Intelligence dapat berkontribusi dalam pengembangan SDM di negara Indonesia. Hasil Pengabdian Pengenalan Artificial Intelligence Kepada Siswa di Lingkungan Sekolah Menengah Atas pada SMA 9 Negeri Samarinda dilakukan selama satu hari yang dihadiri oleh 34 peserta, 2 orang dosen, dan dibantu oleh mahasiswa program studi Ilmu Komputer Universitas Widyagama Samarinda hasil angket dari peyelenggaraan pengabdian ini didapatkan 7 pertanyaan yang hasilnya diatas 80% peserta yang artinya peserta setuju telah mendapat pengetahuan Pengenalan Artificial Intelligence dengan baik atau penyelenggaraan ini sudah dilaksanakan dengan sangat baik sekali.
Ensemble Learning untuk Model Prediksi Risiko Preeklamsia dan Explainable AI Berbasis SHAP Yudhi Fajar Saputra; Milkhatun; Mahmoud Ahmad Al-Khasawneh; Yazeed Al Moaiad; Aldi Bastiatul Fawait; Sitti Rahmah; Zakaria Ahmad Dahlan
METIK Jurnal Vol. 10 No. 1 (2026): METIK Jurnal Issue Published
Publisher : LPPM Universitas Mulia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47002/kp377403

Abstract

Preeclampsia is a pregnancy complication that poses significant risks to both mother and fetus. Early prediction of preeclampsia risk is crucial to improve maternal healthcare outcomes. This study aims to develop a predictive model for preeclampsia risk using ensemble learning approaches and to enhance model interpretability through Explainable Artificial Intelligence (XAI). The dataset consists of 332 pregnant women who received antenatal care, with 330 complete clinical records after data cleaning. Two ensemble learning algorithms, Random Forest (RF) and eXtreme Gradient Boosting (XGBoost), were implemented and evaluated using Receiver Operating Characteristic (ROC) curves, Area Under the Curve (AUC), and additional classification metrics. The best-performing model was further analyzed using SHapley Additive exPlanations (SHAP) to assess feature contributions at both global and individual levels. The results indicate that XGBoost outperformed Random Forest with an AUC of 0.81 compared to 0.72 after applying class weighting and 5-fold cross-validation. XGBoost also demonstrated more balanced performance with an accuracy of 0.83, recall of 0.85, and specificity of 0.60. In contrast, Random Forest achieved an accuracy of 0.91 and specificity of 0.98 but failed to detect positive cases, with a recall of 0.00, indicating bias toward the majority class. SHAP analysis reveals that height, weight, age at menarche, and the number of antenatal care (ANC) visits significantly influence prediction, while hypertension consistently contributes to increased risk. This study demonstrates that integrating ensemble learning with XAI improves both predictive performance and model transparency for preeclampsia risk assessment.
ANALISIS KOMPARATIF KLASIFIKASI MACHINE LEARNING UNTUK MEMPREDIKSI STUNTING PADA ANAK USIA DI BAWAH LIMA TAHUN Yudhi Fajar Saputra; Mahmoud Ahmad Al-Khasawneh; Milkhatun Milkhatun; Ni Wayan Wiwin Asthiningsih; Sitti Rahmah
Syntax : Journal of Software Engineering, Computer Science and Information Technology Vol 6, No 1 (2025): Juni 2025
Publisher : Universitas Dharmawangsa

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.46576/syntax.v6i1.6852

Abstract

Stunting merupakan salah satu permasalahan kesehatan masyarakat yang bisa berdampak jangka panjang terhadap kualitas sumber daya manusia di Indonesia. Deteksi dini terhadap status stunting anak usia di bawah lima tahun menjadi langkah dalam mencegah gangguan pertumbuhan kronis akibat stunting, sehingga penelitian ini bertujuan untuk membangun model klasifikasi status stunting dengan memanfaatkan pendekatan data mining menggunakan algoritma Decision Tree dan Random Forest. Data yang digunakan diperoleh dari hasil survei terhadap ibu yang memiliki anak dibawah umur lima tahun dengan sejumlah 193 responden, data tersebut mencakup variabel antropometri dan sosial ekonomi, seperti tinggi badan, berat badan, usia anak, pendidikan orang tua, pendapatan keluarga, dan urutan kelahiran. data tersebut diproses melalui tahapan Knowledge Discovery in Databases (KDD) meliputi seleksi atribut, imputasi, encoding, dan klasifikasi melalui proses permodelan data mining, selanjutnya evaluasi dilakukan dengan metrik klasifikasi Classification Accuracy(CA) dan Area Under the Curve (AUC) dari kurva Receiver Operating Characteristic (ROC). Hasil penelitian menunjukkan bahwa model Random Forest memiliki performa lebih baik dibandingkan Decision Tree dengan nilai CA 71% dan AUC 0.74. dibandingkan Decision Tree dengan nilai CA 67% dan AUC 0.68. Peneliti berharap bahwa Model prdiksi ini berpotensial dapat digunakan sebagai sistem deteksi dini stunting berbasis data atau sebagai rujukan untuk penelitian berikutnyaKata Kunci—Stunting, Machine Learning, Random Forest, Decision Tree, Classification Model, ROC Curve. ABSTRACTStunting is one of the public health issues that can have long-term impacts on the quality of human resources in Indonesia. Early detection of stunting status among children under five years of age is a critical step in preventing chronic growth disorders. Therefore, this study aims to develop a classification model for stunting status using a data mining approach with Decision Tree and Random Forest algorithms. The dataset was obtained from a survey of 193 mothers with children under five, encompassing anthropometric and socioeconomic variables such as height, weight, child’s age, parental education, family income, and birth order. The data were processed through the stages of Knowledge Discovery in Databases (KDD), including attribute selection, imputation, encoding, and classification modeling. The model performance was evaluated using classification metrics: Classification Accuracy (CA) and the Area Under the Curve (AUC) from the Receiver Operating Characteristic (ROC) curve. The results show that the Random Forest model outperformed the Decision Tree, achieving a CA of 71% and an AUC of 0.74, compared to the Decision Tree with a CA of 67% and an AUC of 0.68. This predictive model is expected to be useful as a data-driven early detection system for stunting or serve as a reference for future research.Keywords—Stunting, Machine Learning, Random Forest, Decision Tree, Classification Model, ROC Curve.
Peningkatan Wawasan Statistika pada Siswa SMK Muhammadiyah Loa Janan Sitti Rahmah; Muh. Jamil; Aldi Bastiatul Fawait; Yudhi Fajar Saputra; Yulindawati
JURPIKAT Vol 7 No 2 (2026)
Publisher : Politeknik Piksi Ganesha Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.37339/jurpikat.v7i2.3097

Abstract

Rendahnya literasi statistika siswa di SMK Muhammadiyah Loa Janan menjadi dasar pelaksanaan program Pengabdian kepada Masyarakat (PKM). Permasalahan ini penting karena penguasaan statistika merupakan bagian dari kompetensi numerasi yang diperlukan oleh lulusan SMK dalam menghadapi dunia kerja serta pengambilan keputusan berbasis data. Berdasarkan hasil observasi dan wawancara dengan guru, sebagian besar siswa masih mengalami kesulitan dalam memahami konsep dasar statistika, seperti pengumpulan, penyajian, dan interpretasi data. Kondisi tersebut dipengaruhi oleh rendahnya kemampuan dasar matematika yang berdampak pada menurunnya motivasi serta kepercayaan diri siswa dalam pembelajaran. Program ini bertujuan untuk meningkatkan literasi statistika siswa melalui pembelajaran interaktif berbasis modul kontekstual serta menilai efektivitasnya. Kegiatan melibatkan 40 siswa kelas XII, satu guru pendamping, empat dosen, dan dua mahasiswa. Tahapan kegiatan meliputi sosialisasi, pelatihan interaktif melalui diskusi, simulasi, latihan soal, serta evaluasi menggunakan post-test. Analisis data menggunakan uji chi-square (goodness of fit). Hasil menunjukkan adanya perbedaan distribusi nilai yang signifikan (χ² hitung = 26,02 > χ² tabel = 11,07; α = 0,05) dengan dominasi kategori rendah dan sedang.
Peningkatan Pemahaman Siswa Sekolah Menengah Kejuruan terhadap Pengolahan Citra Digital dan Computer Vision melalui Kegiatan Edukasi dan Praktik Sederhana Muh Jamil; Sitti Rahmah; Aldi Bastiatul Fawait; Yudhi Fajar Saputra
JURPIKAT Vol 7 No 2 (2026)
Publisher : Politeknik Piksi Ganesha Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.37339/jurpikat.v7i2.3140

Abstract

Pesatnya perkembangan teknologi berbasis Computer Vision membuat berbagai aplikasi pengolahan citra menjadi semakin mudah untuk digunakan oleh siswa. Namun, kemudahan tersebut tidak selalu diikuti dengan pemahaman mengenai bagaimana teknologi tersebut bekerja. Kegiatan pengabdian ini bertujuan untuk memberikan pemahaman dasar mengenai pengolahan citra digital dan Computer Vision kepada siswa sekolah menengah kejuruan. Kegiatan ini dilaksanakan melalui penyampaian materi dan praktik sederhana menggunakan Google Teachable Machine, kemudian dievaluasi melalui pretest dan posttest. Hasil analisis menunjukkan bahwa sebagian besar indikator mengalami peningkatan yang signifikan (p < 0,05). Di sisi lain, siswa juga menunjukkan respon yang sangat positif dengan tingkat minat dan persepsi yang berada pada kategori tinggi hingga sangat tinggi. Meskipun demikian, beberapa konsep dasar seperti pixel dan resolusi masih perlu diperdalam. Namun secara keseluruhan, kegiatan ini mampu meningkatkan pemahaman sekaligus menumbuhkan ketertarikan siswa terhadap teknologi Computer Vision.
Key Factors Associated with Nursing Care Intensity for Nursing Service Management in a Psychiatric Hospital: A Cross-Sectional Study Approach Milkhatun Milkhatun; Yudhi Fajar Saputra; Alfi Ari Fakhrur Rizal; Ni Wayan Wiwin Asthiningsih; Rusni Masnina; Maridi M Dirdjo
Ahmar Metastasis Health Journal Vol. 6 No. 1 (2026): Ahmar Metastasis Health Journal
Publisher : Yayasan Ahmad Mansyur Nasirah

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.53770/amhj.v6i1.789

Abstract

Nursing care intensity is an important indicator of nursing service quality and patient safety. However, in many clinical settings, nursing workload and care demand assessments are still largely based on medical diagnoses rather than psychosocial and cognitive conditions that may better reflect patients’ actual nursing needs. This study examined the association between patient-related factors across four domains (health history and genetic factors, emotional and cognitive functioning, self-concept, and social–interpersonal relationships) and nursing care intensity levels. A cross-sectional study was conducted at RSJD Atma Husada Mahakam Samarinda, Indonesia, using data from 491 patients collected through a standardized nursing care intensity assessment instrument based on the NANDA–NIC–NOC framework. Nursing care intensity was categorized into low, moderate, and high levels, with most patients classified in the moderate category. Associations between variables and nursing care intensity were analyzed using Chi-square tests, and the strength of associations was evaluated using Cramer’s V and Phi coefficient to identify key associated factors. The result showed that most psychosocial, emotional, cognitive, and social variables showed significant associations with nursing care intensity. The strongest associationships were observed for self-esteem (V = 0.487), socialization in the community (V = 0.408), emotion (V = 0.405), perception (V = 0.290), and intelligence (V = 0.280). In contrast, genetic factors and medical diagnosis showed no significant association with nursing care intensity. Nursing care intensity is more strongly associated with psychosocial and cognitive factors than by traditional medical indicators. These findings highlight key associated factors identified through bivariate analysis and suggest that integrating these domains into routine nursing assessments may support evidence-based staffing allocation, reduce missed nursing care, and improve nursing service quality.
Performance Analysis of Ensemble Learning Models Comparing Bagging and Boosting Techniques for Early Preeclampsia Risk Detection in Pregnant Women Prediction Yudhi Saputra; Milkhatun Milkhatun; Aldi Bastiatul Fawait; Zakaria Ahmad Dahlan; Yazeed Al Moaiad; Haviluddin Haviluddin; Rayner Alfred
International Journal of Engineering, Science and Information Technology Vol 6, No 1 (2026)
Publisher : Malikussaleh University, Aceh, Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.52088/ijesty.v6i1.1880

Abstract

Preeclampsia is a major pregnancy complication that substantially contributes to maternal morbidity and mortality worldwide, making early identification of risk factors essential for effective prevention and timely clinical intervention. This study evaluates the performance of ensemble learning models by comparing bagging and boosting techniques to develop an accurate early prediction system for preeclampsia risk using clinical medical record data. The research follows the Cross-Industry Standard Process for Data Mining (CRISP-DM) framework, encompassing data understanding, data preparation, modeling, evaluation, and interpretation. The dataset was obtained from RSUD Inche Abdoel Moeis Samarinda and underwent preprocessing procedures, including data cleaning, transformation, feature encoding, normalization, and Synthetic Minority Over-sampling Technique (SMOTE) to address class imbalance. Six ensemble learning algorithms were evaluated, consisting of Random Forest, Extra Trees, and Rotation Forest as bagging methods, and XGBoost, LightGBM, and CatBoost as boosting methods. Model performance was assessed using accuracy, precision, recall, and weighted F1-score. The experimental results demonstrate that Random Forest achieved the highest predictive performance, with an accuracy of 0.92, precision of 0.93, recall of 0.92, and weighted F1-score of 0.91, indicating superior robustness and generalization capability. Extra Trees achieved comparable accuracy (0.92) but exhibited lower prediction stability across evaluation metrics. Among the boosting algorithms, LightGBM and CatBoost each obtained an accuracy of 0.89, while XGBoost achieved 0.88. Rotation Forest recorded the lowest accuracy (0.62), suggesting limited suitability for this clinical dataset. These findings indicate that bagging-based ensemble methods, particularly Random Forest, outperform boosting techniques for imbalanced clinical data and provide strong empirical support for developing reliable Clinical Decision Support Systems (CDSS) for early preeclampsia screening and risk assessment in healthcare settings
Optimasi Penjadwalan PelajaranDi Hari Selasa SMKN 16 Samarinda dengan Graf Pewarnaan Algoritma Greedy dan Welch Powell Sitti Rahmah; Yulindawati Yulindawati; Muh Jamil; Aldi Bastiatul Fawait; Yudhi Fajar Saputra
DiJITAC : Digital Journal of Information Technology and Communication DiJITAC, Vol 7 No.1, Oktober 2026
Publisher : Universitas Islam Negeri Sultan Aji Muhammad Idris Samarinda

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.21093/dijitac.v7i1.13068

Abstract

Lesson scheduling in schools is often done manually, which is time-consuming and potentially leads to scheduling conflicts between teachers, classes, and subjects. This study aims to apply the concept of Graph Coloring using the Welch-Powell Algorithm in compiling lesson schedules at SMKN 16 Samarinda. The data used include 23 subjects, 28 teachers, and 13 classes in the even semester of the 2025/2026 academic year, focusing on the Tuesday schedule. The research stages start from data collection, graph formation based on teacher-class relationships, adjacency matrix construction, and graph coloring using the Welch-Powell algorithm. Each vertex in the graph represents a teacher, while edges indicate relationships or potential scheduling conflicts. The results show that the graph coloring process produces five different colors representing five groups of time slots without any scheduling conflicts. This indicates that the Welch-Powell method can help the schedule creation process more systematically, efficiently, and structured. Thus, the application of graph coloring using the Welch-Powell algorithm can be an alternative solution in optimizing lesson scheduling in the school environment.