Research Publications

Authors

Azis, Huzain; Salim, Yulita, Department of Informatics Engineering, Universitas Muslim Indonesia, Jl. Urip Sumoharjo No.km.5, Makassar, 90231, Indonesia; Puspitasari, Yustisia, Orthodontic Department, Dentistry Faculty, Universitas Muslim Indonesia, Jl. Urip Sumoharjo No.km.5, Makassar, 90231, Indonesia

Abstract

In this work, we explore the use of deep learning models for soft tissue profile classification in cephalometric images, an essential task in orthodontic treatment planning. To increase classification accuracy and to preprocess the data and remedy class imbalance, four deep learning architectures were used: Convolutional Neural Networks (CNN), Long Short-Term Memory (LSTM), Recurrent Neural Networks (RNN), and Generative Adversarial Networks (GAN). RNN performed best, with an accuracy of 59.76%, followed by LSTM, with 51.22%. CNN showed signs of overfitting, with a validation accuracy of 95.61% but a testing accuracy of 47.56%. Although the GAN model was used to augment the data and reduce class imbalance, it did not improve classification performance. We used evaluation metrics like precision, recall, and F1 Score to assess how well the models performed. Future work will involve larger datasets and advanced regularization techniques to improve generalizability and clinical relevance. Future directions should focus on expanding the dataset to enhance model accuracy and robustness. Additionally, exploring advanced techniques is essential to prevent overfitting. Furthermore, implementing more sophisticated data augmentation methods and continuously optimizing architectures are recommended to achieve improved classification outcomes and support clinical applications for orthodontic tasks.

Related SDGs

Authors

Ayyub, Muhammad Algifari; Salim, Yulita; Mude, Muh. Aliyazid, Universitas Muslim Indonesia

Abstract

Penelitian ini bertujuan untuk membandingkan kinerja antara Chatbot Meta AI dan Grok AI dalam merespons pola perintah yang serupa. Metode yang digunakan meliputi Metode Black Box dan algoritma K-Means Clustering. Hasil pengujian menunjukkan bahwa Grok AI unggul dalam memberikan penjelasan mendalam dan menyajikan jurnal penelitian terkini. Keunggulan ini ditemukan berkat kemampuan pemrosesan bahasa alami yang handal serta integrasi teknologi real-time melalui platform X (Twitter). Di sisi lain, Meta AI menunjukkan keunggulan dalam efisiensi penggunaan, kecepatan respons yang mencapai 11,11 detik, serta kreativitas dalam output, yang didukung oleh optimasi model bahasa Llama 3. 2. Proses clustering dengan K=2 berhasil mengelompokkan data menjadi dua cluster: Cluster 0, di mana Chatbot Grok AI unggul dalam kedalaman informasi yang terkini; dan Cluster 1, di mana Chatbot Meta AI lebih cepat, efisien, serta menawarkan solusi inovatif. Analisis faktor mengungkapkan bahwa perbedaan kinerja ini dipengaruhi oleh desain arsitektur model, strategi integrasi data, serta prioritas pengembangan masing-masing platform

Related SDGs

Authors

Syafie, Lukman; Awangga, Narendra; Salim, Yulita, Universitas Muslim Indonesia

Abstract

Daily rainfall data plays an important role in hydrological and climatological analysis, especially in tropical regions characterised by high rainfall variability and sharp seasonal changes. However, observational data often has gaps, which can reduce model accuracy and obscure relevant climatological signals. This study addresses these issues by applying the Trend-Based Adaptive Window Imputation (TBaWI) method, an adaptive imputation approach that considers local temporal trends and seasonal dynamics in estimating missing rainfall values. This method was tested using CHIRPS data for the Makassar region for the period 2014–2023 with synthetic data loss scenarios of 10%, 15%, 20%, and 25%. The results show that TBaWI consistently provides a lower Mean Absolute Error (MAE) value, namely 6.14–7.65 mm, compared to linear interpolation, which produces 6.46–7.75 mm. The SMAPE value of TBaWI is also lower, for example 33.16% in the 15% data loss scenario, compared to interpolation at 35.06%. In addition, this method showed an improvement in the ability to identify dry days through the Zero Hit Rate (ZHR), which reached 60.08% in the 20% data loss scenario, higher than the interpolation of 58.32%, while the Rainy Hit Rate (RHR) remained in a stable range of 79–88%. These findings indicate that TBaWI is more effective in maintaining climatological consistency and numerical accuracy of tropical rainfall data. Further research is expected to integrate spatial aspects and optimise machine learning-based parameters to improve the generalisation of the method under various climatic conditions.

Related SDGs

Authors

Yunus, Sattar, Department of Environmental Engineering, Faculty of Engineering, Universitas Muslim Indonesia, Indonesia; Marzuki, Ismail, Department of Chemical Engineering, Faculty of Engineering, Universitas Fajar, Indonesia; Hasnawi, Mardiyyah, Department of Computer Science, Faculty of Computer Science, Universitas Muslim Indonesia, Indonesia; Anggamulia, Muh Ilham, Department of Environmental Engineering, Faculty of Engineering, Universitas Muslim Indonesia, Indonesia

Abstract

Atmospheric particulate matter poses significant risks to the environment and human health, depending on its mass concentration and chemical composition. This study aims to quantify the mass concentrations of total suspended particles (TSP) and particulate matter with an aerodynamic diameter of less than 10 µm (PM10), as well as to determine their elemental composition. Particulate samples were collected from three ambient air monitoring sites within the City of Makassar, the capital of South Sulawesi Province, during the peak dry season (May to September 2025). The measured concentrations of PM10 and TSP were evaluated against the Primary Impact Zone (PIA) standards of the Indonesian Air Quality (IAQ) guidelines. A high-volume air sampler (HVAS) was employed to collect particulate samples, and gravimetric analysis was conducted to determine their mass concentrations. The elemental composition of the particles was analyzed using scanning electron microscopy (SEM) coupled with energy-dispersive X-ray spectroscopy (EDS). The mean concentrations of PM10 and TSP were 34.20 µg/m3 and 73.22 µg/m3, respectively. The findings indicate that the mean mass concentrations of both TSP and PM10 were below the Indonesian regulatory standards, although temporal and spatial variations were observed. The identified elemental constituents included Al, Ca, K, Na, Rb, and Sr, suggesting contributions from crustal sources. These results provide valuable insights into the characteristics of particulate matter in residential areas of Makassar and support the development of improved, source-oriented air quality management strategies.

Related SDGs

Authors

Hasnawi, Mursyid, Department of Civil Engineering, Faculty of Engineering, Universitas Muslim Indonesia, Makassar, Indonesia; Hasnawi, Mardiyyah, Department of Computer Science and Information Technology, Universitas Muslim Indonesia, Makassar, Indonesia

Abstract

Reservoir sedimentation critically threatens the long-term sustainability of water infrastructure in tropical monsoon basins. This study presents an integrated assessment of sediment accumulation, morphodynamic evolution, and storage capacity loss in the Bili-Bili Reservoir, Indonesia, based on multi-year bathymetric surveys conducted in 1997 (baseline) to 2024. Multitemporal bathymetric datasets were combined with three-dimensional morphometric reconstruction and exponential predictive modeling (R² = 0.94) to quantify spatiotemporal sediment distribution and forecast future storage decline. The results indicate a 35% reduction in total storage from 375 × 10⁶ m³ at commissioning to 243.85 × 10⁶ m³ in 2024 with projections suggesting a 50% loss by 2048 under current management practices. The 2004 Mount Bawakaraeng landslide contributed an unprecedented sediment inflow of approximately 3.8 × 10⁶ m³ yr⁻¹, while post-2011 sediment-control structures retained approximately 57% of total sediment inflow. Bathymetric evidence further demonstrates progressive upstream deltaic aggradation, reducing hydraulic efficiency and effective storage capacity. Unlike previous short-term or empirical studies, this research introduces a multi-decadal, data-driven framework that explicitly links sedimentation processes with infrastructure performance and management outcomes. The findings provide a quantitative benchmark for reservoir sustainability and inform adaptive sediment management strategies consistent with Indonesia’s IWRM framework. This approach offers a transferable methodology for sediment-prone reservoirs across tropical regions under increasing hydroclimatic variability.

Related SDGs

Authors

Mude, Muh Aliyazid; Umar, Fitriyani, Universitas Muslim Indonesia

Abstract

Masalah efisiensi pada identifikasi kendaraan secara manual telah diidentifikasi sebagai kendala utama dalam sistem keamanan. Sebuah perancangan sistem deteksi pemilik kendaraan berbasis Automatic Number Plate Recognition dan Internet of Things yang mengacu pada standar International Organization for Standardization / International Electrotechnical Commission 30141 telah dikembangkan dalam penelitian ini. Fokus utama riset dilakukan terbatas pada tahap perancangan prototipe menggunakan emulator fritzing guna memvalidasi arsitektur sistem secara virtual. Hasil perancangan menunjukkan bahwa berhasil divisualisasikan dengan tingkat akurasi logika pada 5 domain pada ISO/EIC 130141 yakni device layer, gateway & network, data management, application layer dan domain business layer Disimpulkan bahwa model perancangan pada emulator fritzing ini layak dan dijadikan acuan awal dalam pengembangan sistem keamanan kendaraan sebelum dilakukan implementasi pada perangkat fisik sehingga untuk pengembangan selanjutnya disarankan menggunakan elumator lainya agar ada gambaran yang jelas penggunaan tools berbasis IoT

Related SDGs

Authors

Akbar, Muh Ilham Nur Hidayat; Fattah, Farniwati; Gaffar, Andi Widya Mufila, Universitas Muslim Indonesia

Abstract

This study investigates the classification of Kodály hand signs using hand landmark features extracted from MediaPipe Hands. A landmark-based approach was chosen because it represents hand gestures as structured numerical features that are more efficient than raw image data. The objective of this study is to compare the performance of three tree-based machine learning algorithms, namely Decision Tree, Random Forest, and XGBoost, for classifying eight Kodály hand sign classes (Do–Do’). A dataset consisting of 8,000 samples was collected with a balanced distribution of 1,000 samples per class from 2 participants using the right hand only. Each sample is represented by 63 features derived from 21 hand landmarks with (x, y, z) coordinates. The data were divided using an 80% training and 20% testing holdout scheme and further validated using Stratified 5-fold cross-validation. Model performance was evaluated using accuracy, macro-precision, macro-recall, macro F1-score, and confusion matrix analysis. Experimental results show that XGBoost achieves the best performance with an accuracy of 0.9712 and macro F1-score of 0.9714 in the holdout evaluation and remains stable in cross-validation with an accuracy of 0.9768±0.0046. Random Forest achieved slightly lower performance, while Decision Tree produced the lowest accuracy. Confusion matrix analysis indicates that most misclassifications occur between classes with similar landmark patterns, particularly La and Mi. Therefore, XGBoost is recommended as the most effective model for landmark-based Kodály hand sign classification.

Related SDGs

Authors

Anjasmara, Edo Ranov; Fattah, Farniwati; Gaffar, Andi Widya Mufila, Universitas Muslim Indonesia

Abstract

Waste management, particularly in sorting plastic bottle and non-bottle waste, remains a challenge in supporting effective recycling systems. This study aims to implement the MobileNetV2 model on the ESP32-CAM device using the Edge Impulse platform and to evaluate the model’s performance in real-time object classification. The dataset consists of 200 images containing bottle and non-bottle objects with variations in lighting conditions, shooting angles, and backgrounds. The model was trained using Edge Impulse and then converted into TensorFlow Lite format for deployment on the ESP32-CAM device.The training results show that the model achieved high performance with an accuracy of 92.50%, supported by an AUC of 0.97, precision of 0.98, recall of 0.97, and F1-score of 0.97. Based on the simplified confusion matrix with visual verification, the model achieved 100% accuracy in detecting bottle objects and 95% accuracy for non-bottle objects, with a 5% misclassification rate. However, during real-world implementation on the ESP32-CAM device, the model’s performance decreased to 65.7% accuracy due to differences between training and real-world conditions as well as hardware limitations.Despite this, the system successfully performed real-time image classification on an embedded device. This study demonstrates that the edge artificial intelligence approach using MobileNetV2 and Edge Impulse can be effectively applied to resource-constrained devices, although improvements are still needed in terms of model generalization and dataset diversity.

Related SDGs

Authors

Fattah, Farniwati; Gaffar, Andi Widya Mufila, Universitas Muslim Indonesia

Abstract

Vehicle detection in low-resolution images remains a significant challenge in computer vision, particularly for embedded devices such as ESP32-CAM with limited computational resources and simple image resolution. This study evaluates the performance of YOLOv8 on low-resolution QVGA (320 × 240 pixels) images for vehicle detection and classification. The dataset was independently collected in a controlled laboratory environment using miniature vehicles, covering four vehicle classes (motorcycle, car, bus, and truck) with a total of 4,000 images and a 70:20:10 data split. A pretrained YOLOv8 model was fine tuned for 100 epochs and tested on an ESP32-CAM prototype. The evaluation results demonstrate excellent performance, achieving precision of 0.999, recall of 1.000, mAP@0.5 of 0.995, and mAP@0.5-0.95 of 0.995 on the validation data, as well as real-time detection accuracy of 97% for motorcycles and cars, and 99% for buses and trucks. These findings indicate that YOLOv8 can deliver reliable vehicle detection performance on low-resolution images and is suitable for implementation in embedded device-based systems

Related SDGs

Authors

Zulfikar, Achmad ; Fattah, Farniwati; Gaffar, Andi Widya Mufila, Universitas Muslim Indonesia

Abstract

Introduction: Angklung is a traditional Indonesian musical instrument that continues to evolve through digital technology. However, computer vision–based gesture recognition for controlling physical angklung instruments remains limited. This study investigates landmark-based recognition of Kodály hand signs and evaluates its application for real-time angklung interaction. Method: Hand landmarks were extracted using MediaPipe Hands from RGB camera input. Each gesture was represented by 63 normalized numerical features derived from 21 landmarks. The dataset consists of 8,000 images representing eight Kodály gesture classes (Do–Do'). Gesture classification was performed using the Extreme Gradient Boosting (XGBoost) algorithm. Model evaluation applied a subject-independent two-fold scheme using accuracy, precision, recall, F1-score, and confusion matrix analysis. Real-time system trials were conducted under different lighting conditions and capture distances, and TCP communication with an ESP32 controller was evaluated. Results: The model achieved 96.63% accuracy in Fold 1 and 96.40% in Fold 2. Misclassifications were mainly observed between visually similar gestures, particularly La and Mi. Separate real-time system trials showed consistent recognition under bright lighting, while accuracy decreased under dim lighting, especially for Do (90%) and Mi (86.7%). Gesture recognition remained reliable up to approximately 1.5 m. TCP testing over 200 command events recorded 0% failed acknowledgments with a mean round-trip time of 87.36 ms. Conclusion: These indicate that landmark-based Kodály gesture classification using MediaPipe Hands and XGBoost can support real-time angklung interaction under controlled conditions, although improvements are needed for low-light environments and visually similar gestures

Related SDGs

Authors

Hasanuddin, Tasrif; R, Muhammad Alif Maulana; Darwis, Herdianti; Gaffar, A. Widya Mufila; Satra, Ramdan; Fattah, Farniwati, Universitas Muslim Indonesia

Abstract

The threat of Distributed Denial of Service (DDoS) attacks is a serious challenge to network stability and security, especially in centralized Software Defined Networking (SDN) architectures. This study aims to evaluate and compare the performance of three classification algorithms in detecting DDoS attacks, namely Fully Connected Neural Network (FCNN), Extreme Gradient Boosting (XGBoost), and Random Forest. The dataset used is public SDN network traffic data that has gone through a normalization and feature selection process to improve model efficiency and accuracy. The evaluation was conducted on four scenarios of training and test data sharing ratios (90:10,80:20,70:30, and 60:40) using accuracy, precision, recall, and F 1-score metrics. The test results show that Random Forest obtained the highest accuracy of 99.61 %, followed by FCNN at 94.95 %, and XGBoost at 99.01 %. Despite the slightly lower accuracy, XGBoost showed the best computational efficiency with the fastest training time of 0.94 seconds and the fastest prediction time of 0.03 seconds, making it an ideal choice for real-time detection system implementation. These findings contribute to the development of a precise, efficient, and adaptive DDoS detection system. Future research is recommended to explore advanced feature selection methods such as Recursive Feature Elimination (RFE) to improve the overall performance and efficiency of the model.

Related SDGs

Authors

Fajri; As’ad, Ihwana; Abdullah, Syahrul Mubarak, Universitas Muslim Indonesia

Abstract

Era digital telah membawa angin segar bagi perkembangan global usaha, khususnya pelaku Usaha Mikro, Kecil, serta Menengah (UMKM). Teknologi yang sekarang semakin canggih serta sangat simpel diakses telah sebagai katalisator bagi pertumbuhan dan perkembangan UMKM. Dulu, UMKM mungkin sangat terbatas di pasar lokal dan mempunyai ketergantungan yang sangat tinggi di interaksi secara langsung. tetapi, dengan adanya teknologi UMKM sekarang bisa menjangkau pasar yg jauh lebih luas. menggunakan memanfaatkan teknologi secara efektif UMKM bisa tumbuh serta berkembang pesat. tetapi, UMKM juga perlu terus beradaptasi dengan perkembangan teknologi yg sangat cepat. Penggunaan E-Commercee sangat bermanfaat untuk peningkatan penjualan, memperluas jangkauan pasar, efisiensi operasional, dan penghematan biaya . akan tetapi masih para pelaku perjuangan yg menjalankan penyewaan alat outdoor secara manual sehingga mengakibatkan beberapa kendala mirip berita stok tak sesuai waktu nyata, manajemen pencatatan yang tidak rapi, sebagai akibatnya terjadi resiko double booking. kondisi ini sangat berdampak terhadap turunnya kepuasan pelanggan. oleh sebab itu, sangat diharapkan sebuah solusi digital yang bisa mengatasi pertarungan tadi sekaligus bisa menyampaikan pengalaman kepada pengguna yg optimal. Peneltian ini juga memakai metode Design Thinking yang terdiri asal beberapa tahapan empathize, define, ideate, prototype, dan test. Metode ini dipilih sebab sangat mampu membantu pengembang memahami kebutuhan pengguna secara pribadi serta bisa menciptakan sebuah desain yg sinkron dengan konflik yg terjadi. Selain itu, penelitian ini sangat menekankan pada kualitas desain secara eksklusif dan pengalaman pengguna (UI/UX) buat menaikkan kepuasan serta minat dan loyalitas pelanggan. yang akan terjadi berasal penelitian ini memberikan bahwa perancangan perangkat lunak penyewaan alat outdoor berbasis mobile mampu menjadi suatu solusi yg sangat efektif buat menaikkan efisiensi pengelolaan, memberikan kemudahan buat pengguna dalam melakukan pencarian, pemesanan, dan pembayaran. Fitur tambahan juga berupa integrasi kalender ketersediaan serta metode perhitungan jarak (Haversine) pula menyampaikan nilai lebih, terutama dalam membanatu pelanggan menemukan penyewaan outdoor terdekat. menggunakan demikian, perangkat lunak ini tidak hanya mendukung transformasti digital UMKM, namun pula mampu menaikkan daya saing antar usaha penyewaan alat outdoor pada era digital

Related SDGs

Authors

Akbar, Firman; Purnawansyah; Faradibah, Amaliah, Universitas Muslim Indonesia

Abstract

Penelitian ini bertujuan untuk menganalisis sentimen masyarakat di media sosial X terhadap kebijakan penundaan pengangkatan Calon Pegawai Negeri Sipil (CPNS) dan Pegawai Pemerintah dengan Perjanjian Kerja (PPPK) di Indonesia menggunakan algoritma Naïve Bayes. Data dikumpulkan melalui proses crawling pada media sosial X yang menghasilkan 283 komentar terkait isu tersebut. Proses analisis meliputi pre- processing data (pembersihan, case folding, filtering, stemming, dan tokenizing), pelabelan sentimen positif dan negatif, ekstraksi fitur menggunakan metode TF-IDF, klasifikasi dengan algoritma Multinomial Naïve Bayes, serta evaluasi model menggunakan metrik akurasi, presisi, recall, dan F1-score. Hasil penelitian menunjukkan bahwa algoritma Naïve Bayes mampu melakukan klasifikasi sentimen secara efektif terhadap opini publik yang beragam, mulai dari dukungan, kritik, hingga pandangan netral terhadap kebijakan penundaan tersebut. Temuan ini diharapkan dapat menjadi masukan bagi pemerintah dalam merumuskan kebijakan yang lebih responsif terhadap opini publik

Related SDGs

Authors

Irsyad, Muhammad; Satra, Ramdan; Faradibah, Amaliah, Universitas Muslim Indonesia

Abstract

This study aims to create and implement a network monitoring system with a telegram bot using a public IP on the 2nd floor access point of FIKOM UMI. The method used is the PPDI method. This study aims to monitor the computer network connection system at the nearest access point on the 2nd floor of FIKOM UMI. Currently, the process of monitoring internet network constraints in the laboratory is still being checked manually, so the process is slow and less efficient. The monitoring system uses the Telegram API Bot and L2tp VPN with Mikrotik to obtain information in the form of real-time telegram notification messages to network administrators. The telegram API bot will send notification messages when internet connection problems occur, both down and up. The results of the internet network monitoring system are very helpful for network administrators in searching for information when internet connection problems occur on the 2nd floor of FIKOM UMI

Related SDGs

Authors

Iqbal, Muhammad Farhan ;Fattah, Farniwati, Universitas Muslim Indonesia

Abstract

Seiring dengan perkembangan zaman, maka kebutuhan manusia akan alat pengawasan/pengamanan ikut berkembang, salah satunya contoh yaitu perkebunan. Permasalahan yang sering terjadi pada saat ini adalah kasus gagal panen dikarenakan hama yang seringkali terjadi karena kurangnya pengawasan. Oleh karena itu, karya tulis ini bertujuan untuk merancang sebuah sistem keamanan/pengawasan pada kebun yang dapat dikendalikan secara jarak jauh menggunakan perangkat Smartphone, Sensor PIR, ESP32, Panel Surya, dan Buzzer. Sistem ini dirancang untuk memberikan solusi efektif dalam mengawasi bahkan mengusir hewan penganggu dengan fitur kendali yang fleksibel dan responsif. ESP32 akan bertindak sebagai otak utama sistem, mengumpulkan dan menganalisis data dari Sensor PIR untuk mendeteksi pergerakan. memungkinkan pengguna untuk mengakses sistem melalui perangkat seluler. Sistem ini juga dilengkapi dengan Buzzer untuk memberikan notifikasi audio dalam situasi darurat atau terdeteksi sebuah gerakan dan notifikasi di Aplikasi Telegram. Melalui integrasi komponen-komponen ini, diharapkan sistem dapat memberikan pengawasan yang handal dan memberikan pemilik kebun kendali penuh melalui aplikasi seluler

Related SDGs

Authors

Kaunar, Rifaldi Fitra Akbar; Satra, Ramdan; Fattah, Farniwati, Universitas Muslim Indonesia

Abstract

This research aims to design an Arduino-based Solar Tracking system as a monitoring center, allowing users to monitor panel movements. The method used in this research is an experimental method to determine the performance of the solar tracking tool that has been made, by using a servo motor as a panel driver, LDR sensor and MPU6050 with wireless communication can display information to the user's PC/Smartphone through the blynk application. Where users can monitor panel movements against light in real-time. Based on the test results, the prototype that has been made can be used to monitor solar tracking in real time with an average error of 1.55%.

Related SDGs