Claim Missing Document
Check
Articles

Found 25 Documents
Search

Evaluasi Performa Model Ensemble Learning dalam Deteksi Serangan Jaringan Internet of Things pada Dataset CIC-BCCC-IOT-HCRL-2019 Yudi Raharja; Agung Budi Susanto; Tukiyat Tukiyat
Journal of Informatics and Electronics Engineering Vol. 5 No. 02 (2025): Desember 2025
Publisher : Unit Penelitian dan Pengabdian kepada Masyarakat Politeknik TEDC Bandung Jl. Pesantren Km 2 Cibabat Cimahi Utara – Cimahi 40513 Jawa Barat – Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.70428/jiee.v5i02.1454

Abstract

Perkembangan pesat perangkat Internet of Things (IoT) membawa peningkatan kompleksitas sekaligus risiko pada keamanan jaringan. Studi ini bertujuan untuk menilai performa lima algoritma Ensemble Learning, yaitu Random Forest, AdaBoost, CatBoost, XGBoost, dan LightGBM, dalam Sistem Deteksi Intrusi (IDS) pada jaringan IoT dengan menggunakan dataset CIC-BCCC-IoT-HCRL-2019. Metode penelitian melibatkan tahap pra-pemrosesan data termasuk penerapan dua teknik normalisasi yaitu MinMaxScaler dan Normalizer, serta evaluasi model menggunakan validasi silang 5-Fold Cross-Validation dan pembagian data latih dan uji dengan rasio 80:20. Hasil eksperimen menunjukkan algoritma boosting seperti XGBoost, CatBoost, dan LightGBM secara konsisten memiliki kinerja lebih baik dibandingkan dengan model bagging tradisional seperti Random Forest. XGBoost yang dikombinasikan dengan MinMaxScaler mencapai akurasi tertinggi sebesar 0,9980, sementara LightGBM dengan MinMaxScaler mencatat waktu pelatihan tercepat yakni 2,54 detik. Temuan ini mengindikasikan bahwa penggunaan teknik boosting bersama normalisasi MinMaxScaler dapat secara signifikan meningkatkan akurasi serta efisiensi IDS berbasis IoT. Kata Kunci— Internet of Things, Deteksi Intrusi, Machine Learning, Ensemble Learning, Boosting, Normalisasi Data.
Rainfall Prediction Using a Hybrid CNN-LSTM Method to Support Hydrometeorological Disaster Mitigation in Bima Regency Agus Salim; Tukiyat; Agung Budi Susanto
Jurnal Teknologi Informatika dan Komputer Vol. 12 No. 2 (2026): Jurnal Teknologi Informatika dan Komputer
Publisher : Universitas Mohammad Husni Thamrin

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.37012/jtik.v12i2.3544

Abstract

Rainfall is one of the key meteorological parameters that significantly influences the occurrence of hydrometeorological disasters, such as floods and landslides, particularly in Bima Regency, which has a high level of vulnerability due to climate variability. Accurate rainfall prediction is essential for supporting decision-making in disaster mitigation and water resource management. This study aims to develop a rainfall prediction model using a Hybrid Convolutional Neural Network (CNN)–Long Short-Term Memory (LSTM) approach that effectively captures both the spatial and temporal characteristics of meteorological data. The dataset consists of historical rainfall records along with supporting meteorological variables collected from observation stations over a specified period. The research methodology includes data preprocessing, normalization using the Min-Max Scaling method, time series data construction, development of the Hybrid CNN–LSTM model, model training and testing, and performance evaluation using Mean Absolute Error (MAE), Root Mean Square Error (RMSE), and Mean Absolute Percentage Error (MAPE) metrics. The results indicate that the proposed Hybrid CNN–LSTM model provides accurate rainfall predictions by effectively extracting local features through CNN layers and capturing long-term temporal dependencies through LSTM layers. Furthermore, the proposed model demonstrates more stable performance in predicting rainfall fluctuations compared with conventional approaches, making it a promising solution for supporting early warning systems and decision-making in hydrometeorological disaster mitigation in Bima Regency. Therefore, the application of the Hybrid CNN–LSTM method offers an effective artificial intelligence-based approach to improving rainfall prediction accuracy.
Classification of Low-Level Wind Shear on Imbalanced Radiosonde Data Using a Hybrid CNN-LSTM Architecture Sarnubih Hasan; Agung Budi Susanto; Arya Adhyaksa Waskita
CSRID (Computer Science Research and Its Development Journal) Vol 18 No 3 (2026): Oktober 2026
Publisher : LPPM Universitas Potensi Utama

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.22303/csrid-.18.3.2026.426-438

Abstract

Low-Level Wind Shear (LLWS) during the takeoff and landing phases poses a severe and dynamic threat to aviation safety. However, it remains inherently challenging to detect directly from surface observations due to its highly non-linear, transient, and sequential nature in the lower atmosphere. This study proposes an advanced deep learning approach to classify LLWS events based on high-resolution vertical radiosonde profile data spanning the 2024–2025 period at Soekarno-Hatta International Airport, Indonesia. Addressing the critical challenge of a highly imbalanced dataset—where hazardous meteorological events are exceptionally rare compared to normal atmospheric conditions—the classification target was simplified into an event-level binary scheme. This scheme utilizes a hazard intensity threshold of ≥ 4 knots/100 ft, adhering strictly to the International Civil Aviation Organization (ICAO) moderate-to-severe hazard categories. The comprehensive research methodology involved fixed-grid interpolation into 21 uniform vertical observation points (spanning 0 to 2000 ft), rigorous wind vector transformation into zonal and meridional components, and circular encoding for wind direction to maintain absolute numerical stability during model training. A hybrid Convolutional Neural Network and Long Short-Term Memory (CNN-LSTM) architecture was developed to capture both local spatial gradients and long-term vertical dependencies, which was then systematically compared against a Transformer baseline equipped with a global self-attention mechanism. The empirical results and operational discussion demonstrate that the proposed CNN-LSTM model significantly outperformed the Transformer baseline, achieving an outstanding Area Under the Precision-Recall Curve (PR-AUC) of 0.9672 compared to the Transformer's 0.7533. Furthermore, the confusion matrix analysis conclusively proved the superior capability of the CNN-LSTM network in minimizing false negatives (missed detections) from 17 down to merely 7 critical events. In conclusion, the integration of local feature extraction through CNN and sequential dependency modeling via LSTM is highly effective for addressing imbalanced meteorological datasets, providing a highly robust and physically consistent computational foundation for automated aviation safety early warning systems.
Perfomance Comparison of LSTM and Transformer Model for Rainfall Data Imputation on Automatic Weather Station Tri Nurmayati; Agung Budi Susanto; Arya Adhyaksa Waskita
CSRID (Computer Science Research and Its Development Journal) Vol 18 No 3 (2026): Oktober 2026
Publisher : LPPM Universitas Potensi Utama

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.22303/csrid-.18.3.2026.491-501

Abstract

High-resolution rainfall observations from Automatic Weather Stations (AWS) are essential for hydrometeorological analysis, yet operational sensor and communication failures frequently create missing values. This study compared Long Short-Term Memory (LSTM) and Transformer architectures for imputing missing 10-minute rainfall observations recorded at the Banten Climatological Station AWS in South Tangerang during 2021-2025. Rainfall was modeled using a sliding window containing rainfall at the previous three intervals together with concurrent air temperature and relative humidity. The chronologically ordered dataset comprised 242,497 valid samples and was divided into training, validation, and testing subsets using a 60:20:20 ratio. Each architecture was trained with the Adam optimizer for up to 100 epochs under batch sizes of 32, 16, and 8, with early stopping used to control overfitting. Performance was assessed using root mean square error, mean absolute error, and the coefficient of determination. The best LSTM configuration used a batch size of 8 and achieved an RMSE of 1.1520, an MAE of 0.5140, and an R² of 0.6626 on the test set. The Transformer with a batch size of 8 obtained an RMSE of 0.8287, an MAE of 0.3635, and an R² of 0.8244. These results demonstrate that self-attention represented short-range multivariate dependencies more effectively than recurrent memory under the imbalanced rainfall distribution Keywords :automatic weather station; imputation; rainfall; LSTM; tansformer..
Intelligent DevSecOps Automation Framework Based on AI-Driven Security Analytics for Agent-Based Jenkins Pipelines Risko; Agung Budi Susanto; Arya Adhyaksa Waskita
Jurnal Teknologi Informatika dan Komputer Vol. 12 No. 2 (2026): Jurnal Teknologi Informatika dan Komputer
Publisher : Universitas Mohammad Husni Thamrin

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.37012/jtik.v12i2.3615

Abstract

The adoption of Continuous Integration/Continuous Delivery (CI/CD) has accelerated the implementation of DevSecOps as an approach for integrating security throughout the software development lifecycle. Although various automated security tools are widely available, their outputs remain fragmented, require extensive manual analysis, and provide limited support for security-driven decision-making within CI/CD pipelines. This gap creates the objective of this study, which is to design and evaluate an Intelligent DevSecOps Automation Framework that combines AI-Driven Security Analytics with an Agent-Based Jenkins Pipeline to integrate multiple security tools, normalize heterogeneous findings, enforce a unified security quality gate, and generate contextual security advisories through Large Language Models (LLMs). The study employed an applied research method with a prototype development and experimental approach. The proposed framework integrates Ruff, Semgrep, SonarQube, and Trivy, and is supported by a finding normalization and consolidation mechanism, risk-based prioritization, security quality gate evaluation, and AI-driven advisory generation orchestrated through LangChain. The framework was evaluated using a deliberately vulnerable payment-service repository to examine its capability in detecting vulnerabilities, consolidating cross-tool findings, classifying risk, and producing mitigation recommendations together with a release decision. The results show that the framework successfully unified four security tools within a single pipeline, producing 160 raw findings that were normalized and consolidated into 37 unique findings, of which 27 were categorized as Critical or Major. Fifteen of the highest-priority findings were forwarded to the AI-Driven Security Analytics component, yielding an Overall Risk and Exploitation Risk of Critical, 14 items classified as release blockers, and an overall release recommendation of Hold. These findings indicate that the proposed framework effectively supports early vulnerability detection, cross-tool consolidation, risk-based prioritization, security quality gate enforcement, and AI-assisted security advisory generation to strengthen security decision-making in modern DevSecOps pipelines.