Indonesian Journal of Electrical Engineering and Computer Science
Not a member yet
9109 research outputs found
Sort by
Trust evaluation in online social networks for secured user interactions
Get PDFOnline social network is a good platform, where users can share their opinions, ideas, products, and reviews with known (friends and relatives) and unknown users. The growing fame and its easy accesses of new users sometimes lead to security and privacy issues. Many methods are reported so far to address these issues but usage of high complex cryptographic algorithms creating new set of performance related challenges to the mobile users. In this paper, light weight soft security (trust) method is proposed. The proposed method “Trust evaluation in online social networks for secured user interactions-TEOSN” uses user social activities in estimation of his trustworthiness. Each user is observed in terms of followed factor- (his interactions with others) and follower factor- (others interaction with him). The factors and are estimated using fuzzy logic and user trust- is estimated using beta distribution. The performance of TEOSN is verified theoretically and practically. In experimental results, TEOSN is verified against different number of users; especially it outperformed existing methods in trust computation of target users at 2 to 4-hop distances
Comparative study of deep learning approaches for cucumber disease classification
Get PDFCucumber leaf diseases, such as downy mildew and leaf miner, pose significant challenges to crop yield and quality. Accurate and timely detection is essential to efficient management. The current research assesses seven convolutional neural network (CNN) models for the classification of diseases of cucumber leaves: DenseNet121, InceptionV3, ResNet50V2, VGG16, Xception, MobileNetV2, and NASNet. The dataset includes images from the cucumber disease recognition dataset (Mendeley) and 500 real-time images captured between December 2022 and February 2023 in Karnataka, covering varied lighting conditions. After augmentation, the dataset is divided into testing, validation, and training sets and includes 804 leaf miner, 807 downy mildew, and 804 healthy images. With an overall test accuracy of 99.37% and nearly flawless precision, recall, and F1-scores in every class, ResNet50V2 showed exceptional performance. InceptionV3 and MobileNetV2 also exhibited strong performance with accuracies of 97.29% and 97.70%, respectively. DenseNet121, VGG16, Xception, and NASNet performed well but were slightly outperformed by the top models. The findings indicate ResNet50V2 as the most reliable model for cucumber leaf disease classification, providing a robust foundation for developing automated disease detection systems. This work demonstrates how precise disease detection using deep learning models can improve agricultural management
Deep belief network classification model for accurate breast cancer detection and diagnosis
Get PDFBreast cancer is still one of the common malignancies and endemics that are fatal to women across the globe. Early-stage diagnosis helps reduce the percentage of deaths because treatment outcomes are much better at that stage. As the contemporary approaches in machine learning (ML) and deep learning (DL) emerged, the automatic detection of breast cancer has received a great consideration for their ability to improve diagnosis and treatment. We present a new deep belief network (DBN) based breast cancer detection system to increase the accuracy and the dependability of the diagnosis of breast cancer. The major modules of the system are image preprocessing, feature extraction and the DBN-based classification to guarantee accurate detection and classification of malignant and benign breast lesions. We compared the proposed DBN model with the existing DL models like convolutional neural networks (CNNs), recurrent neural networks (RNNs), long short-term memory (LSTM), and generative adversarial networks (GANs). It is with respect to critical features of the model performance which includes accuracy, precision, recall, specificity and F1-score. The methodologies used in this study show that the performance of the proposed DBN model is significantly better than these conventional algorithms in accuracy and sensitivity where the DBN model is an ideal method for the early detection of breast cancer. Through extensive experimentation, we compared the proposed DBN model with existing DL techniques such as CNNs, RNNs, LSTMs, and GANs. Our results show that the proposed DBN model outperforms these models in several key performance metrics
Real-time driver drowsiness detection based on integrative approach of deep learning and machine learning model
Get PDFDriver drowsiness is a major factor that contributing to road accidents. Several researches are ongoing to detect driver drowsiness, but they suffer from the complexity and cost of the models. This paper introduces a hybrid artificial intelligence (AI)-driven framework integrating deep learning (DL) and machine learning (ML) models for real-time drowsiness detection. The system utilizes a robust DL model to classify driver states based on facial images and support vector machine (SVM) model is trained to develop a cost-efficient yet robust facial landmark detector to extract key features such as eye aspect ratio (EAR) and mouth aspect ratio (MAR). We also introduce a multi-stage decision fusion mechanism that combines convolutional neural network (CNN) probability scores with EAR/MAR thresholds to enhance detection reliability and reduce false positives. Experimental results demonstrate that the proposed model achieves 98% accuracy and F1-score, significantly outperforming traditional DL approaches. Additionally, the SVM-based landmark predictor shows improved efficiency with lower mean squared error (MSE) without having higher computational requirements
Enhancing trust and privacy in iot ecosystems with the distributed trust and privacy consensus framework
Get PDFIn the contemporary digital landscape, the proliferation of wireless sensor networks (WSNs) and the internet of things (IoT) has revolutionized the way we interact with the physical world, offering unprecedented opportunities for automation and data-driven decision-making. However, this rapid expansion has also introduced significant challenges in terms of ensuring network security, maintaining user privacy, and establishing trust among devices. To address these critical issues, this paper introduces the distributed trust and privacy consensus framework (DTPCF), a novel methodology designed to strengthen trust and privacy within IoT ecosystems through a consensus-based approach. The DTPCF pioneers a distributed mechanism for trust management that evaluates and establishes the reliability of nodes democratically and transparently, thereby enhancing the robustness and scalability of IoT systems against malicious activities. Moreover, the framework integrates privacy preservation directly into the consensus process, employing state-of-the-art cryptographic techniques and protocols to protect sensitive data during transmission and decision-making phases. Through empirical analysis, the efficacy of the DTPCF is validated across various operational scenarios, demonstrating its effectiveness in enhancing network security, privacy, and trust. Performance metrics such as throughput, energy consumption, and node-level security are meticulously evaluated, providing comprehensive insights into the framework's capabilities and potential for real-world implementation
S-commerce: competition drives action through small medium enterprise top management
Get PDFThis study investigates the factors influencing the continued use of S-commerce in small and medium enterprises (SMEs), focusing on the roles of top management (TM) support, competitive pressure (CP), facilitating conditions, and service quality. Data were collected from 341 SME owners and analyzed using SEM. Data was analyzed with SmartPLS using a two-step approach. The findings indicate that TM support significantly impacts the continued use of S-commerce by influencing facilitating conditions and service quality while CP affects TM behavior and usage continuity. However, the findings reveal that operational factors, such as infrastructure and service quality, play a more critical role in sustaining S-commerce engagement than external pressures. Facilitating conditions, in particular, were found to have a strong influence on service quality and platform engagement, underscoring the importance of technical and organizational resources. The study extends prior research by highlighting the interplay between internal and external drivers in fostering the continuous use of S-commerce, offering practical insights for SMEs and future research directions
A recurrent network technique for energy optimization in 6G networks with dynamic device-to-device communication
Get PDFEnergy efficiency has become a paramount concern in the design and deployment of 6G networks, driven by the exponential growth of connected devices and increasing traffic demands. For domain experts grappling with dynamic device-to-device (D2D) communication scenarios, optimizing energy consumption while maintaining reliable connectivity poses a significant challenge. To address this issue, we propose a novel recurrent network technique that dynamically configures D2D communication patterns, adaptively allocating temporary base stations among network nodes to enable efficient data transmission while minimizing energy expenditure. Our simulations demonstrate substantial energy savings, extended node lifetimes, and reliable performance, with a 37% reduction in overall network energy consumption and a 65% increase in average node lifetime compared to traditional cellular communication scenarios. In conclusion, this innovative approach paves the way for sustainable and energy efficient 6G communication systems, benefiting society by reducing operational costs, minimizing environmental impact, and prolonging the usability of mobile devices
Enhancing the ternary neural networks with adaptive threshold quantization
Get PDFTernary neural networks (TNNs) with weights constrained to –1, 0, and +1 offer an efficient deep learning solution for low-cost computing platforms such as embedded systems and edge computing devices. These weights are typically obtained by quantizing the real weight during the training process. In this work, we propose an adaptive threshold quantization method that dynamically adjusts the threshold based on the mean of weight distribution. Unlike fixed-threshold approaches, our method recalculates the quantization threshold at each training epoch according to the distribution of real valued synaptic weights. This adaptation significantly enhances both training speed and model accuracy. Experimental results on the MNIST dataset demonstrates a 2.5× reduction in training time compared to conventional methods, with a 2% improvement in recognition accuracy. On Google Speech Command dataset, the proposed method achieves an 8% improvement in recognition accuracy and a 50% reduction in training time, compared to fixed-threshold quantization. These results highlight the effectiveness of adaptive quantization in improving the efficiency of TNNs, making them well-suited for deployment on resource constrained edge devices
Automatic wildlife species identification on camera trap images using deep learning approaches: a systematic review
Get PDFThe foundation of systematic research depends on precise species identification, functioning as a critical component in the processes of biological research. Wildlife biologists are prompting for more effective techniques to fulfill the expanding need for species identification. The rise in open source image data showing animal species, captured by digital cameras and other digital methods of collecting data, has been monumental. This rapid expansion of animal image data, integrated with state-of-the-art machine learning techniques such as deep learning which has shown significant capabilities for automating species identification. This paper focuses on the role of deep neural network architectures in furthering technological advancements in automating species identification in recent years. To advocate further investigation in this field, an examination of machine learning architectures for species identification was presented in this work. This examination focuses primarily on image analyses and discusses their significance in wildlife conservation. Fundamentally, the aim of this article is to offer insights into the present advancements in automating species identification and to act as a reference for scholars who are keen to integrate machine learning techniques into ecological studies. Systems designed through Artificial Intelligence are extensive in providing toolkits for systematic identification of species in the upcoming years
Intrusion detection system using hybrid CNN-LSTM model in cloud computing
Get PDFCloud computing has revolutionized online service delivery with its flexibility and cost efficiency. Nevertheless, the growing importance of stored data makes it a target for cyberattacks, posing security and privacy risks. This calls for effective solutions to safeguard data and infrastructure, particularly with regard to intrusion attacks and distributed attacks such as distributed denial of service (DDoS). Therefore, there is a need to develop an effective intrusion detection system (IDS) using deep learning to ensure the protection of cloud data and infrastructure. In this paper, a hybrid model aims to leverage the power of convolutional neural networks (CNNs) to analyze spatial features and extract complex patterns, while long short-term memory LSTMs are used to understand temporal data sequences and detect attacks that evolve over time to detect intrusions in cloud computing environments on the CSE-CIC-IDS2018 dataset. The model was trained and tested on DDoS attacks, and the results demonstrated high performance in detecting attacks with high accuracy and efficiency. This hybrid model achieved an accuracy of 99.88%, a precision of 99.83%, a recall of 99.94%, and an F1-score of 99.88%