Abstract:Abstract: The growing intensity of cyber attacks, marked by rapid, large-scale, automated, and adaptive execution, requires analytical methods that represent the diversity of network environments, including variations in…
target platforms such as IoT, traditional networks, and hybrid infrastructures. This study compares machine learning models for cyber attack classification under heterogeneous environmental conditions and formulates a conceptual optimization framework based on model performance. Four publicly available benchmark datasets were used, namely UNB CIC IoT 2023, UNB CIC IDS-2018, UNSW-NB15, and a Kaggle cyber security attacks dataset, comprising approximately 40,000 to over 3.6 million records and 25 to 80 features across IoT, conventional, and mixed network environments. Random Forest, XGBoost, Multilayer Perceptron, and Transformer were implemented within a unified pipeline involving preprocessing, feature selection, and Bayesian Optimization-based hyperparameter tuning. All models achieved F1-score and Cohen's Kappa above 96%, with XGBoost performing best (97.80%, 97.26%), followed by Random Forest (97.78%, 96.96%) and Transformer (97.44%, 96.82%), while MLP scored lowest (96.74%, 96.00%), a gap below one percentage point. Confusion matrix analysis revealed persistent misclassification in minority and overlapping attack classes, informing a proposed adaptive cyber attack simulation optimization framework.
Keywords: cyber attacks; optimization; machine learning; environmental variability.
Abstrak: Meningkatnya intensitas serangan siber yang berlangsung cepat, masif, otomatis, dan adaptif menuntut pendekatan analitis yang merepresentasikan keragaman lingkungan jaringan, termasuk perbedaan karakteristik platform sasaran seperti Internet of Things (IoT), jaringan konvensional, dan infrastruktur hibrida. Penelitian ini membandingkan model machine learning untuk klasifikasi serangan siber pada kondisi lingkungan heterogen, sekaligus menyusun kerangka optimasi konseptual berdasarkan performa model. Empat dataset benchmark publik digunakan, yaitu UNB CIC IoT 2023, UNB CIC IDS-2018, UNSW-NB15, serta dataset Kaggle cyber security attacks, dengan jumlah data berkisar 40.000 hingga lebih dari 3,6 juta rekaman dan 25 sampai 80 fitur, mewakili lingkungan IoT, konvensional, dan campuran. Random Forest, XGBoost, Multilayer Perceptron, dan Transformer diimplementasikan melalui pipeline terpadu mencakup pra-pemrosesan, seleksi fitur, dan optimasi hyperparameter berbasis Bayesian Optimization. Seluruh model mencapai F1-score dan Cohen's Kappa di atas 96%, dengan XGBoost menunjukkan performa terbaik (97,80%, 97,26%), diikuti Random Forest (97,78%, 96,96%) dan Transformer (97,44%, 96,82%), sementara MLP mencatat skor terendah (96,74%, 96,00%), dengan selisih kurang dari satu poin persentase. Analisis confusion matrix mengungkap misklasifikasi yang konsisten pada kelas minoritas dan serangan dengan karakteristik serupa, yang menjadi dasar kerangka optimasi simulasi serangan siber adaptif yang diusulkan.
Kata kunci: serangan siber; optimasi; machine learning; variabilitas lingkungan
Abstract:Abstract: Human metapneumovirus (HMPV) poses a global health threat, but its detection remains challenging due to limited environmental monitoring. This study aims to develop a portable diagnostic tool for rapid HMPV detection…
ection by integrating cutting-edge biotechnology (CRISPR-Cas system and immunoassay) with air quality sensors on an Internet of Things (IoT)-based microfluidic platform controlled by an ESP32 microcontroller. The system is supported by a companion application and data analysis using Vertex AI, and is capable of providing results in less than fifteen minutes. The development results demonstrate the potential for improving detection accuracy and reliability, particularly with further development of virus-specific biosensors, sensor optimization, and algorithms. This technology is effective as a complementary tool for early screening and environment-based risk management in areas with limited laboratory facilities, although it does not completely replace molecular diagnostic methods such as PCR. A rapid diagnostic approach based on environmental sensors, IoT, and artificial intelligence is a promising strategy to improve early HMPV detection, accelerate public health responses, and strengthen respiratory infection prevention through integrated environmental monitoring and education functions.
Keywords: air quality; CRISPR-Cas; Human Metapneumovirus (HMPV); Internet of Things, portable diagnostic; public health; rapid detection; sensors.
Abstrak: Human metapneumovirus (HMPV) merupakan ancaman bagi kesehatan global, namun pendeteksiannya masih sulit akibat keterbatasan pemantauan lingkungan. Studi ini bertujuan mengembangkan alat diagnostik portabel untuk deteksi cepat HMPV melalui integrasi bioteknologi mutakhir (sistem CRISPR-Cas dan immunoassay) dengan sensor kualitas udara pada platform mikrofluida berbasis Internet of Things (IoT) yang dikendalikan mikrokontroler ESP32. Sistem ini didukung aplikasi pendamping dan analisis data menggunakan Vertex AI, serta mampu memberikan hasil dalam waktu kurang dari lima belas menit. Hasil pengembangan menunjukkan potensi peningkatan akurasi dan keandalan deteksi, terutama dengan pengembangan lanjutan berupa biosensor spesifik virus, optimalisasi sensor, dan algoritma. Teknologi ini efektif sebagai alat pelengkap untuk skrining awal dan manajemen risiko berbasis lingkungan di wilayah dengan keterbatasan fasilitas laboratorium, meskipun tidak sepenuhnya menggantikan metode diagnostik molekuler seperti PCR. Pendekatan diagnostik cepat berbasis sensor lingkungan, IoT, dan kecerdasan buatan menjadi strategi menjanjikan untuk meningkatkan deteksi dini HMPV, mempercepat respons kesehatan masyarakat, serta memperkuat pencegahan infeksi saluran pernapasan melalui fungsi pemantauan dan edukasi lingkungan yang terintegrasi.
Kata kunci: CRISPR-Cas; diagnostik portabel; deteksi cepat; Human Metapneumovirus (HMPV); kesehatan masyarakat; IoT (Internet of Things); sensor kualitas udara.
Abstract:Abstract: The rapid development of digital technology has brought significant changes in consumption patterns and customer behavior, particularly in the fashion retail sector. Increasing competition and rising cynsumer expectations…
xpectations for fast, accurate, and technology-based services require businesses to innovate and undergo digital transformation. One widely used approach is the implementation of automation-based Electronic Customer Relationship Management (E-CRM). This study aims to analyze the implementation of E-CRM automation and its impact on service efficiency at Toko Diah Fashion, a fashion retail business that still faces service challenges due to manual systems, such as customer data duplication, delayed responses, and difficulty in monitoring transaction history. The research method used is a descriptive qualitative method with data collection techniques through observation, interviews, and documentation. The research focus is limited to aspects of service efficiency, including service speed, accuracy in managing customer data, and ease in customer follow-up. The E-CRM automation system studied is designed using PHP programming language and a MySQL database. The research results indicate that the implementation of E-CRM automation can significantly improve service efficiency. This system facilitates integrated customer data management, speeds up the service process, and supports more personalized communication through notification features and transaction history recording.
Keyword: automation; e-crm; fashion retail; service efficiency.
Abstrak: Perkembangan teknologi digital yang semakin pesat telah membawa perubahan signifikan dalam pola konsumsi dan perilaku pelanggan, khususnya dalam sektor ritel fashion. Persaingan yang semakin ketat serta meningkatnya ekspektasi konsumen terhadap layanan yang cepat, akurat dan berbasis teknologi menuntut pelaku usaha untuk melakukan inovasi dan transformasi digital. Salah satu pendekatan yang banyak digunakan adalah penerapan Electronic Customer Relationship Management (E-CRM) berbasis automasi. Penelitian ini bertujuan untuk menganalisis penerapan automasi E-CRM serta pengaruhnya terhadap efisiensi pelayanan pada Toko Diah Fashion, sebuah usaha ritel fashion yang masih menghadapi kendala pelayanan akibat sistem manual, seperti duplikasi data pelanggan, keterlambatan respons, dan kesulitan dalam pemantauan histori transaksi. Metode penelitian yang digunakan adalah metode kualitatif deskriptif dengan teknik pengumpulan data melalui observasi, wawancara, dan dokumentasi. Fokus penelitian dibatasi pada aspek efisiensi pelayanan, meliputi kecepatan pelayanan, ketepatan pengelolaan data pelanggan, serta kemudahan dalam tindak lanjut pelanggan. Sistem automasi E-CRM yang dikaji dirancang menggunakan bahasa pemrograman PHP dan basis data MySQL. Hasil penelitian menunjukkan bahwa penerapan automasi E-CRM mampu meningkatkan efisiensi pelayanan secara signifikan. Sistem ini mempermudah pengelolaan data pelanggan secara terintegrasi, mempercepat proses pelayanan, serta mendukung komunikasi yang lebih personal melalui fitur notifikasi dan pencatatan histori transaksi.
Kata kunci: automasi; e-crm; efisiensi pelayanan; ritel fashion.
Abstract:Abstract: Stunting is a chronic nutritional condition in toddlers characterized by a Height-for-Age (HFA) measurement below the standard growth threshold, necessitating early detection to prevent long-term consequences.…
This study aims to classify toddler stunting status by comparing three machine learning methods: Random Forest (RF), K-Nearest Neighbor (KNN), and Support Vector Machine (SVM). The dataset comprises 345 toddler records from Puskesmas Indramayu (2025), including weight, height, and nutritional status based on WFA, HFA, and WFH indicators. Preprocessing steps include data cleaning, StandardScaler normalization, One-Hot Encoding for categorical features, and splitting the training and testing data with a ratio of 80:20. The comparison results are that KNN achieved the best performance with an accuracy of 71.01%, a precision of 0.69, a recall of 0.69, and an F1 score of 0.67, while RF and SVM both had an accuracy of 69.57% with F1 scores of 0.67 and 0.68, respectively. Thus, KNN demonstrated superior effectiveness in classifying the stunting status of toddlers compared to RF and SVM on this dataset.
Keywords: KNN; Random Forest; SVM; Stunting; toddlers
Abstract: Stunting adalah kondisi gizi kronis pada balita yang ditandai dengan pengukuran Tinggi Badan menurut Usia (HFA) di bawah ambang batas pertumbuhan standar, sehingga memerlukan deteksi dini untuk mencegah konsekuensi jangka panjang. Penelitian ini bertujuan untuk mengklasfikasikan status stunting pada balita dengan membandingkan tiga metode pembelajaran mesin: Random Forest (RF), K-Nearest Neighbor (KNN), dan Support Vector Machine (SVM). Kumpulan data terdiri dari 345 catatan balita dari puskesmas indramayu (2025), termaksut brat badan, tinggi badan, dan status gizi berdasarkan indicator WFA, HFA, dan WFH. Langkah-langkah prapemrosesan meliputi pembersian data, normalisasi Stand-ardScaler, One-Hot Encoding untuk fitur kategirikal, serta pembagian data pelatihan dan pengujian dengan rasio 80:20. Hasil perbadingan adalah KNN mencapai kinerja terbaik dengan akurasi 71,01%, presisi 0,69, recall 0,69, dan skor F1 sebesar 0,67, RF dan SVM keduanya memiliki akurasi 69,57% dengan skor F1 masing-masing sebesar 0,67 dan 0,68. Dengan demikian, KNN menunjukkan keefektifan yang lebih unggul dalam mengklasifikasikan status stunting balita dibandingkan dengan RF dan SVM pada da-taset ini.
Kata kunci: KNN; random forest; SVM; Stunting; Balita
Abstract:Abstract: The materials sector is one of the stock markets sectors that attracts investors due to the high level of construction activity in Indonesia, which supports long-term growth. Stock price movements are influenced…
d by various factors, requiring investors to determine the appropriate timing for buying, selling, or holding stocks. Therefore, this study aims to predict stock prices in the materials sector using a combination of CNN–BiLSTM algorithms. The research data were obtained from Yahoo Finance and processed through min–max normalization, data splitting, sliding window, model implementation, and evaluation stages. Testing was conducted on INTP and SMGR stocks with data split scenarios ranging from 60:40 to 90:10. The results show that CNN–BiLSTM performs best with a 90:10 data split, with minimum MSE and MAPE values of 0.000153 and 2.471% for INTP, and 0.000199 and 2.208% for SMGR, respectively. These findings indicate that increasing the proportion of training data improves the model's ability to learn historical patterns and produce more stable predictions.
Keywords: CNN-BILSTM; materials sector; stock
Abstrak: Sektor materials merupakan salah satu sektor saham yang diminati investor karena tingginya aktivitas pembangunan di Indonesia yang mendorong pertumbuhan jangka panjang. Pergerakan harga saham dipengaruhi oleh berbagai faktor sehingga investor perlu menentukan waktu transaksi yang tepat. Oleh karena itu, penelitian ini bertujuan memprediksi harga saham sektor materials menggunakan kombinasi algoritma CNN–BiLSTM. Data penelitian diperoleh dari Yahoo Finance dan diproses melalui tahapan normalisasi min–max, pembagian data, sliding window, implementasi model, serta evaluasi. Pengujian dilakukan pada saham INTP dan SMGR dengan skenario pembagian data 60:40 hingga 90:10. Hasil menunjukkan bahwa CNN–BiLSTM menghasilkan performa terbaik pada pembagian data 90:10, dengan nilai MSE dan MAPE minimum masing-masing sebesar 0.000153 dan 2.471% untuk INTP, serta 0.000199 dan 2.208% untuk SMGR. Temuan ini mengindikasikan bahwa peningkatan porsi data latih meningkatkan kemampuan model dalam mempelajari pola historis dan menghasilkan prediksi yang lebih stabil.
Kata kunci: CNN-BILSTM; saham; sektor materials
Abstract:Abstract: Malaria remains a major global health problem, particularly in tropical countries such as Indonesia. Accurate early diagnosis is essential for reducing malaria-related morbidity and mortality. Conventional microscopic…
oscopic examination is time-consuming, highly dependent on expert personnel, and prone to human error. This study compares the performance of two Convolutional Neural Network (CNN) architectures, ResNet-50 and DenseNet-121, for malaria image classification. The Cell Images for Malaria dataset provided by the National Institutes of Health (NIH) through Kaggle was used, consisting of 27,558 microscopic blood cell images categorized into Parasitized and Uninfected classes. The dataset was divided into 80% training data and 20% testing data. Image preprocessing included resizing to 224 × 224 pixels, normalization, labeling, and data augmentation using RandomFlip, RandomRotation, RandomZoom, and RandomContrast. Experimental results showed that the ResNet-50 model trained for 100 epochs achieved the highest performance, with an accuracy of 95.54% and precision, recall, and F1-score of 0.96. The confusion matrix indicated 5,272 correctly classified images out of 5,510 testing samples. These findings demonstrate that ResNet-50 outperformed DenseNet-121 and has strong potential for supporting accurate, reliable, and efficient computer-aided malaria diagnosis based on microscopic blood smear images.
Keywords: computer-aided diagnosis; convolutional neural network (CNN); densenet-121; early detection; image classification; malaria; microscopic blood smear images; resnet-50;
Abstrak : Malaria masih menjadi masalah kesehatan global yang serius, terutama di negara tropis seperti Indonesia. Diagnosis dini yang akurat sangat penting untuk menurunkan angka morbiditas dan mortalitas. Metode konvensional berupa pemeriksaan mikroskopis memiliki keterbatasan karena memerlukan waktu yang relatif lama, bergantung pada tenaga ahli, dan berpotensi menimbulkan kesalahan manusia. Penelitian ini bertujuan membandingkan kinerja arsitektur Convolutional Neural Network (CNN) yaitu ResNet-50 dan DenseNet-121 dalam klasifikasi citra malaria. Dataset yang digunakan berasal dari Cell Images for Malaria yang disediakan oleh National Institutes of Health (NIH) melalui platform Kaggle, terdiri dari 27.558 citra dengan pembagian 80% data latih, 20% data validasi. Tahap praproses meliputi cleaning, resizing citra menjadi 224×224 piksel, normalisasi, labeling, serta data augmentasi menggunakan RandomFlip, RandomRotation, RandomZoom, dan RandomContrast. Hasil pengujian menunjukkan bahwa model ResNet-50 pada epoch 100 memperoleh akurasi sebesar 95,54% dengan nilai precision, recall, dan F1-score masing-masing sebesar 0,96. Confusion matrix menunjukkan jumlah prediksi benar sebanyak 5.272 dari total 5.510 data uji. Hasil ini menunjukkan bahwa arsitektur CNN mampu mengklasifikasikan citra malaria dengan tingkat akurasi yang tinggi dan memiliki kemampuan generalisasi yang baik terhadap data baru. Penelitian ini memberikan kontribusi dalam evaluasi performa arsitektur CNN untuk mendukung pengembangan sistem diagnosis malaria berbasis citra mikroskopis yang lebih cepat dan akurat.
Kata kunci: convolutional neural network (CNN); citra mikroskopis hapusan darah; densenet-121; diagnosis berbantuan komputer; deteksi dini; klasifikasi citra; malaria; resnet-50
Abstract:This study aims to apply the Analytic Network Process (ANP) method as a decision support tool in determining the eligibility of education grant recipients in North Sumatra Province. The background of this research arises…
from the large number of grant applicants compared to the available budget, as well as the absence of clear and objective evaluation standards. The ANP method was chosen because it allows the interdependence between assessment criteria such as institutional feasibility, performance and achievement, social and educational impact, and accountability and transparency to be analyzed comprehensively. Data were obtained through interviews, documentation, and observation at the North Sumatra Provincial Education Office. The results of the ANP model show that the criterion with the highest weight is accountability and transparency (0.44), followed by social and educational impact (0.31). Among the three alternatives, community-based education foundations (A2) obtained the highest total weight (0.30), indicating that they are the most eligible recipients of education grants. The implementation of the ANP-based decision support system produces valid and consistent ranking results (CR < 0.1), enabling faster, fairer, and more transparent decision-making. Therefore, the ANP method contributes significantly to improving governance, objectivity, and accountability in the distribution of education grants in North Sumatra Province.
Abstract:Abstract: The rapid degradation of mangrove ecosystems threatens coastal biodiversity, shoreline stability, and carbon sequestration capacity, particularly in areas experiencing intense human activity. However, community-based…
-based participatory mangrove monitoring remains limited due to the lack of accessible and user-friendly digital tools. This study aims to design an intuitive mobile application for mangrove tree detection and participatory ecological monitoring using a User-Centered Design (UCD) approach. The research was conducted iteratively through user needs analysis, prototype development, and usability evaluation involving local governments, conservation practitioners, and non-expert users. The proposed application integrates machine learning for automated mangrove recognition with geospatial visualization and real-time feedback to support field-based monitoring. Usability evaluation using the System Usability Scale (SUS) yielded an overall score of 82.3, categorized as excellent usability, indicating high user satisfaction and intuitive interaction. The results demonstrate that integrating UCD and machine learning enhances usability, user engagement, and the accuracy of mangrove documentation under real field conditions. Overall, this study presents a field-ready, user-centered mobile solution that bridges usability engineering and participatory mangrove monitoring as a replicable model for inclusive ecological application development.
Keywords: Carbon sequestration; mangrove monitoring; mobile application; user-centered design; usability evaluation
Abstrak: Degradasi ekosistem mangrove yang semakin cepat mengancam keanekaragaman hayati pesisir, stabilitas garis pantai, dan kapasitas sekuestrasi karbon, terutama di wilayah dengan aktivitas manusia yang intens. Namun, pemantauan mangrove secara partisipatif berbasis komunitas masih terbatas akibat kurangnya perangkat digital yang mudah diakses dan ramah pengguna. Penelitian ini bertujuan merancang aplikasi mobile yang intuitif untuk deteksi pohon mangrove dan pemantauan ekologi partisipatif dengan menggunakan pendekatan User-Centered Design (UCD). Penelitian dilakukan secara iteratif melalui analisis kebutuhan pengguna, pengembangan prototipe, dan evaluasi kegunaan dengan melibatkan pemerintah daerah, praktisi konservasi, serta pengguna non-ahli. Aplikasi yang diusulkan mengintegrasikan pembelajaran mesin untuk pengenalan mangrove secara otomatis dengan visualisasi geospasial dan umpan balik waktu nyata guna mendukung pemantauan di lapangan. Evaluasi kegunaan menggunakan System Usability Scale (SUS) menghasilkan skor keseluruhan sebesar 82,3 yang termasuk dalam kategori kegunaan sangat baik, menunjukkan tingkat kepuasan pengguna yang tinggi dan interaksi yang intuitif. Hasil penelitian menunjukkan bahwa integrasi UCD dan pembelajaran mesin meningkatkan kegunaan, keterlibatan pengguna, serta akurasi dokumentasi mangrove dalam kondisi lapangan. Secara keseluruhan, penelitian ini menyajikan solusi mobile berbasis UCD yang siap digunakan di lapangan dan menjembatani rekayasa kegunaan dengan pemantauan mangrove partisipatif sebagai model replikatif bagi pengembangan aplikasi ekologi yang inklusif.
Kata kunci: Carbon sequestration; mangrove monitoring; mobile application; user-centered design; usability evaluation
Abstract:Abstract: Mental health is an essential aspect of overall well-being, particularly for university students vulnerable to emotional strain. This study aims to identify clusters of student mental health trends using the K-Means…
Means clustering technique. The research involved 60 students from four academic programs at the Faculty of Science and Technology, selected using stratified and cluster sampling techniques. Data were collected using a modified Mental Health Inventory (MHI). The results revealed distinct commonalities among majors: the Statistics program was predominantly defined by the depressed cluster at 53.3%, while Mathematics followed at 40% within the same cluster. In contrast, Biology students predominantly fell under the neu-tral/stable cluster (66.7%), whilst Information Systems students exhibited an even distribution (33.3% per cluster) without a dominant trend. The clustering quality was evaluated using the Silhouette Coefficient, yielding a range of 0.39 to 0.60. Biology (0.60) and Statistics (0.54) exhibited a reasonable structure, but Information Systems (0.39) and Mathematics (0.34) demonstrated a deficient structure. In conclusion, K-Means effectively discerns mental health patterns, providing a data-driven basis for targeted psychological interventions in educational settings.
Keywords: biology; information systems; k-means; mathematics; mental health; silhouette coefficient; statistics
Abstrak: Kesehatan mental merupakan komponen vital dari kesejahteraan total, terutama bagi maha-siswa yang rentan terhadap stres emosional. Penelitian ini bertujuan untuk mengidentifikasi kelompok tren kesehatan mental mahasiswa melalui penerapan metode pengelompokan K-Means. Studi ini mencakup 60 mahasiswa dari empat program studi di Fakultas Sains dan Teknologi, yang dipilih melalui metode pengambilan sampel bertingkat dan kelompok. Data dikumpulkan dengan menggunakan Inventaris Kesehatan Mental (MHI) yang dimodifikasi. Temuan menunjukkan kesamaan yang jelas di antara jurusan: program studi Statistika terutama ditandai oleh kelompok depresi (53,3%), diikuti oleh Matematika dengan 40% dalam kelompok depresi. Sebaliknya, mahasiswa Biologi terutama termasuk dalam kelompok netral/stabil (66,7%), sedangkan mahasiswa Sistem Informasi memiliki distribusi yang merata (33,3% per kelompok) tanpa pola yang dominan. Kualitas pengelompokan dinilai dengan Koefisien Sil-houette, menghasilkan rentang 0,39 hingga 0,60. Biologi (0,60) dan Statistika (0,54) memiliki struktur sedang, sedangkan Sistem Informasi (0,39) dan Matematika (0,34) menunjukkan struktur yang buruk. Kesimpulannya, K-Means secara akurat mengidentifikasi tren kesehatan mental, menawarkan landasan berbasis data untuk terapi psikologis yang ditargetkan di ling-kungan pendidikan.
Kata kunci: biologi; kesehatan mental; K-Means; matematika; silhouette coefficient; sistem in-formasi; statistika
Abstract:Abstract: Accurate bone age estimation is essential for monitoring pediatric growth, diagnosing endocrine disorders, and supporting clinical decision-making. Although deep learning has improved prediction accuracy, limited…
ed studies have systematically examined how increasing model depth affects performance and reliability. This study evaluates the effectiveness of progressively deeper convolutional neural networks, specifically EfficientNet variants B0 to B5, for bone age estimation from hand radiographs. Experiments were conducted using 12,611 hand X-ray images from the RSNA Pediatric Bone Age Challenge dataset on Kaggle. To ensure fair comparison, all models were trained using a unified and consistent training pipeline. Model performance was evaluated using Mean Absolute Error (MAE), Mean Absolute Percentage Error (MAPE), Concordance Correlation Coefficient (CCC), and Pearson correlation coefficient. The results show a consistent improvement in prediction accuracy as model depth increases. Among the evaluated models, EfficientNet-B5 achieved the best performance, with an MAE of 21.5 months, MAPE of 6.23%, CCC of 0.9148, and Pearson’s r of 0.9203. These findings confirm that model scaling plays a critical role in enhancing prediction robustness and clinical reliability. Future work should emphasize external validation across diverse populations and incorporate interpretability techniques, such as Grad-CAM, to improve clinical transparency and trust.
Keywords: bone age prediction; deep learning; model evaluation; clinical validation
Abstrak: Estimasi usia tulang yang akurat sangat penting untuk memantau pertumbuhan anak, mendiagnosis gangguan endokrin, dan mendukung pengambilan keputusan klinis. Meskipun pembelajaran mendalam telah meningkatkan akurasi prediksi, studi yang secara sistematis meneliti bagaimana peningkatan kedalaman model memengaruhi kinerja dan keandalan masih terbatas. Studi ini mengevaluasi efektivitas jaringan saraf konvolusional yang semakin dalam, khususnya varian EfficientNet B0 hingga B5, untuk estimasi usia tulang dari radiografi tangan. Eksperimen dilakukan menggunakan 12.611 gambar sinar-X tangan dari dataset RSNA Pediatric Bone Age Challenge di Kaggle. Untuk memastikan perbandingan yang adil, semua model dilatih menggunakan alur pelatihan yang terpadu dan konsisten. Kinerja model dievaluasi menggunakan Mean Absolute Error (MAE), Mean Absolute Percentage Error (MAPE), Concordance Correlation Coefficient (CCC), dan koefisien korelasi Pearson. Hasil menunjukkan peningkatan yang konsisten dalam akurasi prediksi seiring dengan peningkatan kedalaman model. Di antara model yang dievaluasi, EfficientNet-B5 mencapai kinerja terbaik, dengan MAE sebesar 21,5 bulan, MAPE sebesar 6,23%, CCC sebesar 0,9148, dan Pearson’s r sebesar 0,9203. Temuan ini menegaskan bahwa penskalaan model memainkan peran penting dalam meningkatkan optimasi prediksi dan keandalan klinis. Penelitian selanjutnya dapat menekankan validasi eksternal di berbagai populasi dan menggabungkan teknik interpretasi, seperti Grad-CAM, untuk meningkatkan transparansi dan kepercayaan klinis.
Kata kunci: prediksi usia tulang; deep learning; evaluasi model; validasi klinis