Publication Search

Indonesia di kawasan Cincin Api Pasifik, yang dikenal memiliki aktivitas seismik yang sangat tinggi dengan ribuan gempa bumi yang terjadi setiap tahunnya. Penelitian ini bertujuan untuk menganalisis kinerja Recurrent Neural Network (RNN) dan Long Short-Term Memory (LSTM) dalam memprediksi magnitudo gempa bumi menggunakan data historis yang diambil dari Kaggle. Data tersebut mencakup rentang waktu dari November 2008 hingga September 2022, yang telah melalui proses normalisasi serta perpecahan menjadi data pelatihan dan pengujian. Model evaluasi kinerja dilakukan dengan menggunakan metrik Mean Absolute Error (MAE) dan Root Mean Square Error (RMSE). Pada uji coba pertama, LSTM menunjukkan performa terbaik dengan nilai MAE 0.6226 dan RMSE 0.7731 pada data pengujian, lebih baik dibandingkan RNN yang mencatatkan MAE 0.6271 dan RMSE 0.7831. Sebaliknya, pada uji coba kedua, RNN unggul dengan nilai MAE 0.5583 dan RMSE 0.7008, sementara LSTM memiliki MAE 0.5822 dan RMSE 0.7132. Hasil ini menunjukkan bahwa LSTM lebih cocok untuk menangani pola data temporal yang kompleks, sedangkan RNN lebih andal pada dataset dengan pola yang lebih sederhana. Penelitian ini diharapkan dapat menjadi pijakan dalam pengembangan sistem prediktif untuk mitigasi risiko bencana gempa bumi di Indonesia.

The Industrial Internet of Things (IIoT) enhances the connectivity and efficiency of living lifestyles. However, it also comes with significant security vulnerabilities. Traditional authentication methods are often inadequate, leading to IIoT devices opened to security threats. This paper proposes a comprehensive security framework integrating blockchain, cryptographic techniques, smart contracts, and deep learning-based Intrusion Detection Systems (IDS) to tackle the mentioned issue. Blockchain ensures data integrity and prevents tampering through a decentralized ledger. A decentralized device identity management system enhances user verification, while secure communication protocols using Hash-based Message Authentication Codes (HMAC) safeguard data integrity. Smart contracts automate transactions, providing transparent, secure record-keeping without a central authority. The deep learning-based IDS, utilizing Contractive Sparse Autoencoder (CSAE) and Attention-Based Bidirectional Long Short-Term Memory (ABiLSTM) networks, effectively detects cyber threats. Evaluation metrics, including precision, recall, F1-score, and False Acceptance Rate (FAR), demonstrate high accuracy and low false alarm rates across datasets. This framework addresses the need for secure, efficient, and scalable authentication in IIoT, combining blockchain's security features with advanced cryptographic and anomaly detection techniques, offering robust defence against cyber threats.

The rapid advancement of deep learning technologies has significantly transformed healthcare analytics, particularly in medical data prediction and classification. This study proposes a hybrid Convolutional Neural Network–Long Short-Term Memory (CNN–LSTM) framework for multi-modal healthcare data analysis, integrating medical imaging, structured electronic health records (EHRs), and IoT-generated time-series physiological signals. The proposed architecture combines spatial feature extraction through CNN with temporal dependency modeling via LSTM to enhance predictive accuracy and clinical decision support. A quantitative experimental design was employed, utilizing multi-source healthcare datasets that underwent preprocessing, normalization, and feature engineering prior to model training. The performance of the hybrid model was evaluated using Accuracy, Precision, Recall, F1-Score, AUC-ROC, and Mean Absolute Error (MAE), and compared with conventional machine learning models and standalone deep learning architectures. Experimental results demonstrate that the proposed CNN–LSTM model achieves superior performance, with improved classification accuracy and reduced prediction error, while maintaining strong generalization capability. The findings indicate that integrating spatial and temporal feature learning significantly enhances disease detection, risk stratification, and personalized treatment planning. This approach supports the development of intelligent clinical decision support systems and scalable smart healthcare environments. The proposed framework offers a reliable and efficient solution for advanced healthcare analytics in IoT-enabled systems.

The rapid growth of digital technologies has significantly increased the complexity and frequency of cyber threats, making network security a critical concern in modern information systems. Traditional security approaches, such as rule-based and signature-based systems, are often limited in detecting sophisticated and unknown attacks. Therefore, this study proposes an Anomaly-Based Intrusion Detection System (AbIDS) utilizing machine learning and deep learning techniques to enhance detection capabilities. The research adopts a Design Science Research approach, involving stages of problem identification, data collection, preprocessing, model development, system implementation, and evaluation. Several models, including Decision Tree (DT), Support Vector Machine (SVM), Convolutional Neural Network (CNN), and Long Short-Term Memory (LSTM), are implemented and compared. The results indicate that deep learning models, particularly LSTM and CNN, outperform traditional machine learning methods in terms of accuracy, precision, recall, and F1-score, while maintaining a lower false positive rate. Additionally, the integration of incremental learning enables the system to adapt to new attack patterns without requiring complete retraining, improving scalability and real-time performance. Despite the promising results, challenges such as computational complexity and false positives remain. Overall, the proposed IDS model demonstrates strong potential as an effective and adaptive solution for enhancing network security in dynamic environments.

The rapid growth of social media platforms has generated massive volumes of unstructured textual data containing valuable information about public opinions and sentiments. Extracting meaningful insights from this data has become increasingly important for decision-making in various domains, including business, politics, and social analysis. This study aims to evaluate the effectiveness of deep learning techniques for sentiment analysis of social media data, focusing on Convolutional Neural Networks (CNN), Long Short-Term Memory (LSTM), and a hybrid CNN-LSTM model. A quantitative experimental approach is employed, where datasets are preprocessed through text cleaning, tokenization, and feature representation using word embeddings. The models are trained and evaluated using standard performance metrics, including accuracy, precision, recall, and F1-score. The results indicate that all models perform effectively in sentiment classification tasks, with the hybrid CNN-LSTM model achieving the highest performance due to its ability to capture both local textual features and long-term contextual dependencies. This demonstrates that combining CNN and LSTM architectures enhances classification accuracy compared to individual models. Furthermore, the findings confirm that deep learning approaches are more robust in handling the complexity and noisiness of social media data compared to traditional methods. This study contributes to the development of more adaptive and accurate sentiment analysis models and highlights the potential of hybrid deep learning architectures for real-world applications.

This study compares different machine learning models for time series forecasting in financial data analysis. Models including ARIMA, LSTM, and GRU are applied to predict stock price movements. We measure the accuracy and computational efficiency of each model on various datasets and discuss their strengths and weaknesses in financial forecasting contexts. The findings suggest that deep learning models show significant improvement in capturing complex temporal patterns over traditional methods.

Cyber Physical Systems (CPS) are vital for managing and controlling critical infrastructures, such as industrial control systems, power grids, and transportation networks. These systems integrate digital and physical components, offering numerous benefits for industrial automation. However, the increasing interconnectivity of these systems has introduced new security vulnerabilities, particularly in anomaly detection and system reliability. This research aims to address these challenges by proposing an edge based anomaly detection framework that leverages lightweight deep learning models, specifically designed to operate efficiently on resource constrained edge devices. Literature Review: Previous studies have shown the effectiveness of anomaly detection in CPS, with traditional methods struggling to keep up with the complexity and scale of modern industrial environments. Machine learning and deep learning approaches, particularly hybrid models combining rule based systems and AI, have emerged as effective solutions for real time anomaly detection. Techniques such as model compression, quantization, and pruning are essential for adapting these models to resource limited edge devices while maintaining high detection accuracy and low latency. Materials and Method: The proposed framework integrates deep learning models such as Convolutional Neural Networks (CNNs) and Long Short Term Memory (LSTM) networks, optimized for edge computing environments. The datasets used for training and testing include industrial network traffic and sensor anomaly datasets. Model optimization techniques like pruning and quantization were applied to reduce computational overhead and energy consumption on edge devices. Results and Discussion: The framework demonstrated high detection accuracy (AUC of 0.9720) with ultra low latency (0.0019 seconds training time), making it highly suitable for real time anomaly detection in CPS. Resource efficiency was achieved by optimizing the models for edge devices, reducing energy consumption while maintaining performance. The framework also significantly improved security by identifying anomalies early, preventing potential threats to critical infrastructures. Future directions include exploring federated learning to enhance privacy and data sharing across distributed devices.

Background: Public opinion analysis has become increasingly important in the digital era, where social media platforms generate large-scale textual data reflecting public perceptions toward environmental policies. Advances in Natural language processing (NLP) and machine learning enable systematic sentiment classification to support data-driven decision-making. Objective: This study aims to evaluate the effectiveness of several sentiment classification models in analyzing Indonesian-language social media data related to environmental policies. Method: The research employed a text mining pipeline including data crawling, preprocessing (case folding, tokenization, stopword removal, and stemming), and vectorization using TF-IDF. Three classification models Logistic Regression, Support Vector Machine (SVM), and Long Short-Term Memory (LSTM) were trained and evaluated using accuracy and F1-score metrics. Results: Experimental findings indicate that LSTM achieved the highest performance with 91.7% accuracy and 91.2% F1-score, outperforming SVM (88.5%) and Logistic Regression (84.2%). Sentiment distribution analysis shows that public opinion is dominated by positive sentiment (47.5%), followed by neutral (32.0%) and negative (20.5%). Overall: The results demonstrate that deep learning-based models provide more robust contextual understanding and more reliable sentiment mapping for environmental policy analysis.

Music genre classification is one part of the music recommendation process, which is a challenging job. This research proposes the classification of music genres using Bidirectional Long Short-Term Memory (BiLSTM) and Mel-Frequency Cepstral Coefficients (MFCC) extraction features. This method was tested on the GTZAN and ISMIR2004 datasets, specifically on the IS-MIR2004 dataset, a duration cutting operation was carried out, which was only taken from seconds 31 to 60 so that it had the same duration as GTZAN, namely 30 seconds. Preprocessing operations by removing silent parts and stretching are also performed at the preprocessing stage to obtain normalized input. Based on the test results, the performance of the proposed method is able to produce accuracy on testing data of 93.10% for GTZAN and 93.69% for the ISMIR2004 dataset.