Jurnal Komtika (Komputasi dan Informatika)
Aims Jurnal Komtika (Komputasi dan Informatika) is a scientific journal published by the Faculty of Engineering, Universitas Muhammadiyah Magelang and is Accredited by the Ministry for Research, Technology, and Higher Education (RISTEKDIKTI)(No:200/M/KPT/2020). It is a medium for researchers, academics, and practitioners interested in Computer Science and wish to channel their thoughts and findings. Our concept of Informatics includes technologies of information and communication as well as results of research, critical, and comprehensive scientific study which are relevant and current issues covered by the journals. Jurnal Komtika publishes regular research articles. We encourage researchers to publish their theoretical and empirical results in as much detail as possible. For theoretical papers, full details of proofs must be provided so that the results can be checked; for experimental papers, full experimental details must be given so that the results can be reproduced. Additionally, electronic files or software regarding the full details of the calculations, experimental procedure, etc., can be deposited along with the publication as “Supplementary Material”. Scope Jurnal Komputasi dan Informatika (Komtika) focuses on various issues, but not limited in the field of: Software Development: Software development process, Requirements analysis, Software design, Software construction, Software deployment, Software maintenance, Programming team, Open-source model Mathematics of Computing: Discrete mathematics, Mathematical software, Information theory Theory of computation: Model of computation, Computational complexity Human Computer Interaction: Interaction design, Social computing, Ubiquitous computing, Visualization, Accessibility, User Interface Study, User Experience Study Applied Computing: E-commerce, Enterprise software, Electronic publishing, Cyberwarfare, Electronic voting, Video game, Word processing, Operations research, Educational technology, Document management. Machine Learning: upervised learning, Unsupervised learning, Reinforcement learning, Multi-task learning Graphics: Animation, Rendering, Image manipulation, Graphics processing unit, Mixed reality, Virtual reality, Image compression, Solid modeling Information System: Database management system, Information storage systems, Enterprise information system, Social information systems, Geographic information system, Decision support system, Process control system, Multimedia information system, Data mining, Digital library, Computing platform, Digital marketing, World Wide Web, Information retrieval
Articles
150 Documents
<![CDATA[Monitoring dan Pemberian Pakan Ikan Lele Otomatis berbasis Internet of Things (IoT) di Tambak Good's Lele ]]>
<![CDATA[Putra, Nyoman Adi Andrian Kusuma]]>;
<![CDATA[Paramartha Putra, Made Adi]]>;
<![CDATA[Noviyanti Kusuma, Ni Putu]]>
<![CDATA[ Jurnal Komtika (Komputasi dan Informatika) Vol 9 No 2 (2025) ]]>
Publisher : <![CDATA[Universitas Muhammadiyah Magelang ]]>
Show Abstract
|
Download Original
|
Original Source
|
Check in Google Scholar
|
DOI: 10.31603/komtika.v9i2.15171
<![CDATA[Budidaya ikan lele merupakan sektor potensial dalam memenuhi kebutuhan konsumsi masyarakat. Namun, metode pemberian pakan manual sering menyebabkan ketidakteraturan dan memicu kanibalisme, yang menurunkan produktivitas. Tambak Good’s Lele di Batubulan, Sukawati, Gianyar, masih menggunakan metode manual sehingga diperlukan sistem otomatis untuk meningkatkan efisiensi. Pengembangan sistem ini memanfaatkan microcontroller ESP32 dan dilengkapi dengan berbagai sensor seperti sensor suhu (DS18B20), sensor pH, turbidity sensor, ultrasonic, dan loadcell. Sistem ini mampu memantau kualitas air serta mendeteksi tinggi dan berat pakan dalam wadah. Ketika kondisi terdeteksi sesuai, mekanisme pemberian pakan akan diaktifkan secara otomatis menggunakan motor servo dan motor DC. Data hasil pemantauan ditampilkan melalui LCD 20x4 I2C serta dikirimkan ke antarmuka website yang dapat diakses melalui perangkat seperti laptop atau smartphone. Hasil akhir dari proyek ini adalah sebuah sistem yang terintegrasi dan dapat bekerja secara otomatis serta manual melalui antarmuka website. Sistem ini memungkinkan pengawasan dan pemberian pakan ikan secara tepat waktu dan efisien. Selain itu, sistem ini juga diharapkan dapat membantu meningkatkan produktivitas tambak dan mendukung pengembangan teknologi di sektor perikanan berbasis IoT]]>
<![CDATA[Implementasi Algoritma Random Forest Berbasis Machine Learning Untuk Prediksi Klon Kopi Unggul ]]>
<![CDATA[Febriansyah, Febriansyah]]>;
<![CDATA[Nurmaleni, Nurmaleni]]>
<![CDATA[ Jurnal Komtika (Komputasi dan Informatika) Vol 9 No 2 (2025) ]]>
Publisher : <![CDATA[Universitas Muhammadiyah Magelang ]]>
Show Abstract
|
Download Original
|
Original Source
|
Check in Google Scholar
|
DOI: 10.31603/komtika.v9i2.15227
<![CDATA[The significant increase in coffee prices in recent years has not been matched by optimized production, particularly in major coffee-producing regions such as Pagar Alam City. One of the main challenges is farmers’ limited capacity to determine the most suitable coffee clone for their environmental conditions. This study aims to develop an intelligent system based on machine learning to predict superior coffee clones that can improve productivity and support food security. The Random Forest algorithm was applied using the CRISP-DM framework, consisting of business understanding, data understanding, data preparation, modelling, evaluation, and deployment stages. The dataset comprised environmental variables (altitude, rainfall, soil pH, soil type, pest resistance, and production) and coffee clone labels (Clone1–Clone4). Experimental results indicate that the model achieved an average accuracy of approximately 75% under 5-fold cross-validation, with altitude and rainfall identified as the most influential factors in clone selection. The predictive system was implemented in Python and can be further developed into web- or mobile-based applications. This study demonstrates the potential of artificial intelligence in optimizing coffee production, enhancing farmers’ welfare. Kenaikan harga kopi yang signifikan dalam beberapa tahun terakhir belum diimbangi dengan produksi yang optimal, terutama di wilayah penghasil kopi utama seperti Kota Pagar Alam. Salah satu tantangan utama adalah keterbatasan kemampuan petani dalam menentukan klon kopi yang paling sesuai dengan kondisi lingkungannya. Penelitian ini bertujuan untuk mengembangkan sistem cerdas berbasis machine learning guna memprediksi klon kopi unggul yang dapat meningkatkan produktivitas dan mendukung ketahanan pangan. Algoritma Random Forest diterapkan dengan menggunakan kerangka kerja CRISP-DM, yang meliputi tahap pemahaman bisnis, pemahaman data, persiapan data, pemodelan, evaluasi, dan penerapan. Dataset yang digunakan mencakup variabel lingkungan seperti ketinggian, curah hujan, pH tanah, jenis tanah, ketahanan terhadap hama, serta data produksi, dengan label klon kopi (klon1–klon4). Hasil eksperimen menunjukkan bahwa model yang dibangun mencapai rata-rata akurasi sekitar 75% menggunakan metode 5-fold cross-validation, dengan ketinggian dan curah hujan teridentifikasi sebagai faktor paling berpengaruh dalam pemilihan klon. Sistem prediksi ini diimplementasikan menggunakan Python dan dapat dikembangkan lebih lanjut menjadi aplikasi berbasis web atau mobile. Penelitian ini menunjukkan potensi kecerdasan buatan dalam mengoptimalkan produksi kopi, meningkatkan kesejahteraan petani, serta memperkuat ketahanan pangan nasional.]]>
<![CDATA[Optimasi Gradient Boosted Trees dalam Memprediksi Minat Nasabah untuk Berlangganan Pinjaman Berjangka ]]>
<![CDATA[Achmad, Refi Riduan]]>;
<![CDATA[Zulfariansyah, Muhammad]]>
<![CDATA[ Jurnal Komtika (Komputasi dan Informatika) Vol 9 No 2 (2025) ]]>
Publisher : <![CDATA[Universitas Muhammadiyah Magelang ]]>
Show Abstract
|
Download Original
|
Original Source
|
Check in Google Scholar
|
DOI: 10.31603/komtika.v9i2.15259
<![CDATA[This research focuses on optimizing the design parameters of Gradient Boosted Trees (GBT) to predict customer interest in subscribing to term loans. The study highlights the importance of tuning parameters such as the number of trees, tree depth, and learning rate to enhance the predictive accuracy of GBT. Through this optimization, the model aims to provide more precise insights into customer behavior, aiding financial institutions in making informed decisions and improving operational efficiency. The research compares GBT with other algorithms like Decision Trees and Random Forests, utilizing metrics such as accuracy, precision, recall, and AUC. The results indicate that GBT, with optimal parameter settings, outperforms the other models in predicting customer interest. The study concludes that GBT is an effective tool for market segmentation and can significantly contribute to more accurate predictions in financial services, ultimately helping companies develop better-targeted marketing strategies.]]>
<![CDATA[Segmentasi Berbasis Data Time Series Penjualan Produk Kopi Menggunakan Algoritma K-Means ]]>
<![CDATA[Anggaini, Meri]]>;
<![CDATA[Herlawati, Herlawati]]>;
<![CDATA[Purnomo, Rakhmat]]>
<![CDATA[ Jurnal Komtika (Komputasi dan Informatika) Vol 9 No 2 (2025) ]]>
Publisher : <![CDATA[Universitas Muhammadiyah Magelang ]]>
Show Abstract
|
Download Original
|
Original Source
|
Check in Google Scholar
|
DOI: 10.31603/komtika.v9i2.15336
<![CDATA[Coffee shops are businesses in the Food and Beverage (F&B) sector that contribute 7.15% to Indonesia's economy. The high demand for coffee has led to increasingly fierce competition. Kanae Coffee & Space in Bekasi faces challenges in maintaining customer loyalty and managing unpredictable demand. This study aims to apply the K-Means algorithm to cluster coffee products based on time series sales data, using the 6-step CRISP-DM methodology. The number of clusters was determined using the elbow method and confirmed with a silhouette coefficient of 0.5916 (good structure). The analysis resulted in five clusters with distinct characteristics: Cluster 0 (very low demand, stable trend, very high price), Cluster 1 (very high demand but sharply declining trend, very low price), Cluster 2 (moderately high demand, moderately stable trend, moderate price), Cluster 3 (moderate demand, slowly declining trend, moderately high price), and Cluster 4 (low demand, stable trend, moderately low price). These segmentation results are expected to serve as the basis for more effective marketing strategies and product management.]]>
<![CDATA[Pengelompokkan Abstrak Jurnal Ilmiah Menggunakan Term Frequency-Inverse Document Frequency dan K-Means ]]>
<![CDATA[Nadia Wati Aprianti]]>;
<![CDATA[Herman Yuliansyah]]>;
<![CDATA[Muhammad Kunta Biddinika]]>
<![CDATA[ Jurnal Komtika (Komputasi dan Informatika) Vol. 10 No. 1 (2026) ]]>
Publisher : <![CDATA[Universitas Muhammadiyah Magelang ]]>
Show Abstract
|
Download Original
|
Original Source
|
Check in Google Scholar
|
DOI: 10.31603/komtika.v10i1.15516
<![CDATA[ The rapid growth of scientific publications in Indonesia has created a need for text analysis methods capable of automatically clustering articles based on content similarity and research themes. This study aims to implement a combination of Term Frequency Inverse Document Frequency (TF-IDF) and the K-Means in the process of grouping scientific journal abstracts in the field of informatics. The research data consist of 1,200 scientific journal abstracts manually collected from the official SINTA (Science and Technology Index) portal for the 2023”“2024 publication period, covering various levels of national journal accreditation. The study employs an unsupervised machine learning approach consisting of several stages, including text preprocessing, TF-IDF weighting, clustering using K-Means, and result evaluation using the Silhouette Score and Davies”“Bouldin Index (DBI) metrics. The TF-IDF weighting process produced 3,000 of the most informative terms, dominated by keywords such as data, method, result, and system, reflecting the research characteristics in the field of informatics. The clustering process generated four main clusters with a Silhouette Score of 0.0121 and a DBI value of 8.3996, indicating that the model was able to identify initial thematic similarities among abstracts. The Word Cloud visualization revealed variations in research topic focus across clusters, including algorithm testing, data model development, system applications, and methodological implementation. This study contributes to the development of a national framework for scientific text analysis that can be utilized for research topic mapping, inter-institutional collaboration, and data-driven research policy formulation.]]>
<![CDATA[Pengembangan Prototype Kotak Obat Otomatis Berbasis IoT untuk Pemantauan dan Pengingat Konsumsi Obat ]]>
<![CDATA[Putu Dendy Kayoana Radika]]>;
<![CDATA[Made Adi Paramartha Putra]]>;
<![CDATA[Putu Trisna Hady Permana]]>
<![CDATA[ Jurnal Komtika (Komputasi dan Informatika) Vol. 10 No. 1 (2026) ]]>
Publisher : <![CDATA[Universitas Muhammadiyah Magelang ]]>
Show Abstract
|
Download Original
|
Original Source
|
Check in Google Scholar
|
DOI: 10.31603/komtika.v10i1.16096
<![CDATA[The smart medicine box is a modern solution designed to assist users in managing their medication schedules consistently and on time. This study aims to design an automated reminder and medicine dispensing system based on the Internet of Things (IoT) using an ESP32 microcontroller. The system is connected to a Firebase Realtime Database to store and display data in real-time through a web interface built with Next.js. The device is equipped with a stepper motor, I2C LCD, buzzer, and a manual button that functions as a mechanism for refilling medicine. Testing was conducted using the black-box method to ensure all features functioned properly. The implementation results show that the system successfully performs medicine time reminders, automatic dispensing, and history logging. It is expected that this system can be further developed to enhance user convenience, improve adherence to medication schedules, and ensure reliable remote monitoring through the website platform.]]>
<![CDATA[Analisis Dampak SMOTE terhadap Feature Importance pada Klasifikasi Data Migraine menggunakan Random Forest dan Extra Trees ]]>
<![CDATA[Henny Leidiyana]]>
<![CDATA[ Jurnal Komtika (Komputasi dan Informatika) Vol. 10 No. 1 (2026) ]]>
Publisher : <![CDATA[Universitas Muhammadiyah Magelang ]]>
Show Abstract
|
Download Original
|
Original Source
|
Check in Google Scholar
|
DOI: 10.31603/komtika.v10.i1.16684
<![CDATA[This study analyzes the impact of the Synthetic Minority Over-sampling Technique (SMOTE) on model performance and feature importance in the classification of migraine patients using Random Forest (RF) and Extra Trees (ET) algorithms. Evaluation was conducted based on recall and F1-Score for the minority class, as well as Permutation Importance analysis. The results indicate that ET, especially when combined with SMOTE (ET + SMOTE), delivers the best performance for the minority class. ET + SMOTE achieved an average F1-Score of 0.7000 and an average recall of 0.8041 using 11 optimal features, indicating better feature efficiency. The application of SMOTE significantly affected the ranking of important features. Although SMOTE improved detection for some minority classes, its impact was not always consistent and occasionally reduced performance on other minority classes. This study concludes that SMOTE alters feature contributions and model interpretability, as well as enhances performance on certain minority classes, particularly when combined with ET.]]>
<![CDATA[Perbandingan XGBoost, Random Forest, dan MLP untuk Klasifikasi Kesiapan Atlet ]]>
<![CDATA[Bintang Putra Wardana]]>;
<![CDATA[Cahyono Budy Santoso]]>
<![CDATA[ Jurnal Komtika (Komputasi dan Informatika) Vol. 10 No. 1 (2026) ]]>
Publisher : <![CDATA[Universitas Muhammadiyah Magelang ]]>
Show Abstract
|
Download Original
|
Original Source
|
Check in Google Scholar
|
DOI: 10.31603/komtika.v10.i1.16850
<![CDATA[Athlete readiness classification is critical for optimizing training load and preventing overtraining-related injuries. This study develops and compares three machine learning algorithms XGBoost, Random Forest, and MLP Neural Network to classify athlete readiness into three ordinal categories: Lower, Middle, and Upper. The dataset comprises 1153 instances with nine multidimensional features encompassing physiological parameters (training duration, intensity, interval days, points, recovery) and psychological indicators (mental score, athlete category, consistency). Data preprocessing involved label encoding for categorical variables and standard scaling for numerical features, followed by a stratified 80:20 train-test split. Model performance was evaluated using weighted precision, recall, F1-score, confusion matrix, and one-vs-rest ROC-AUC curves with 5-fold cross-validation. Results indicate that XGBoost achieved the highest predictive performance (F1-score: 0.93, AUC: 0.99), followed by Random Forest (F1-score: 0.91, AUC: 0.98) and MLP Neural Network (F1-score: 0.85, AUC: 0.94). Feature importance analysis revealed that mental score, training intensity, and consistency were the strongest predictors of readiness status. The proposed framework offers a robust, data-driven decision support tool for sports practitioners, enabling objective readiness monitoring and dynamic training adjustments. Future work will focus on real-time wearable integration and automated hyperparameter optimization]]>
<![CDATA[Perbandingan Kinerja Algoritma Machine Learning pada Sentimen #KaburAjaDulu dengan Penanganan Ketidakseimbangan Data Menggunakan SMOTE ]]>
<![CDATA[Alisha Sumahesa]]>;
<![CDATA[Cahyono Budy Santoso]]>
<![CDATA[ Jurnal Komtika (Komputasi dan Informatika) Vol. 10 No. 1 (2026) ]]>
Publisher : <![CDATA[Universitas Muhammadiyah Magelang ]]>
Show Abstract
|
Download Original
|
Original Source
|
Check in Google Scholar
|
DOI: 10.31603/komtika.v10.i1.16851
<![CDATA[The phenomenon of labor migration abroad has become a widely discussed social issue on social media, particularly through the hashtag #KaburAjaDulu on the X platform. This study aims to analyze public sentiment toward the phenomenon using machine learning classification methods with the implementation of Synthetic Minority Over-sampling Technique (SMOTE) to address data imbalance. The research was conducted using 1,750 data collected from the X platform through several stages, including data collection, text preprocessing, sentiment labeling, TF-IDF weighting, SMOTE implementation, and classification using Support Vector Machine (SVM), Naive Bayes, Random Forest, and Logistic Regression algorithms. Model evaluation was carried out using accuracy, precision, recall, and f1-score metrics. The results show that the implementation of SMOTE significantly improved classification performance. Logistic Regression achieved the best performance with an accuracy of 91.36%, followed by Random Forest at 90.30%, Support Vector Machine at 88.89%, and Naive Bayes at 82.89%. These findings indicate that Logistic Regression has the best capability in recognizing sentiment patterns within unstructured social media data. This study proves that data balancing using SMOTE plays an important role in improving sentiment classification performance and in understanding public opinion regarding social phenomena developing in digital media.]]>
<![CDATA[Komparasi Kinerja ResNet50 dan MobileNetV2 pada Klasifikasi Citra Multi-Domain ]]>
<![CDATA[Herlawati Herlawati]]>;
<![CDATA[Andy Achmad Hendharsetiawan]]>;
<![CDATA[Wowon Priatna]]>;
<![CDATA[Afina Putri Dzulqiyana]]>;
<![CDATA[Regita Ari Rahmadanti]]>
<![CDATA[ Jurnal Komtika (Komputasi dan Informatika) Vol. 10 No. 1 (2026) ]]>
Publisher : <![CDATA[Universitas Muhammadiyah Magelang ]]>
Show Abstract
|
Download Original
|
Original Source
|
Check in Google Scholar
|
DOI: 10.31603/komtika.v10.i1.16926
<![CDATA[This study aims to compare the performance of the Convolutional Neural Network (CNN) architectures ResNet50 and MobileNetV2 in multi-domain image classification, namely herbal leaf images and facial skin undertone images. The herbal plant dataset consists of 11 classes of leaf images, while the skin undertone dataset includes warm, cool, and neutral categories obtained from secondary and primary data sources. The research stages include image preprocessing, model training using transfer learning, and performance evaluation using accuracy, precision, recall, F1-score, and confusion matrix. The results show that ResNet50 achieved better performance than MobileNetV2 in both classification domains. In herbal plant classification, ResNet50 achieved an accuracy of 95.00%, while MobileNetV2 obtained 94.09%. In facial skin undertone classification, ResNet50 achieved an accuracy of 89.3% and a weighted F1-score of 0.893, whereas MobileNetV2 achieved an accuracy of 68.3% and a weighted F1-score of 0.684. These findings indicate that ResNet50 is more effective and stable than MobileNetV2 for multi-domain image classification.]]>
