International Journal of Electrical and Computer Engineering (IJECE)
Not a member yet
6355 research outputs found
Sort by
Credit card fraud data analysis using proposed sampling algorithm and deep ensemble learning
Get PDFCredit card fraud detection is challenging due to the severe imbalance between legitimate and fraudulent transactions, which hinders accurate fraud identification. To address this, we propose a deep learning-based ensemble model integrated with a proposed sampling algorithm based on random oversampling. Unlike traditional methods, the proposed sampling algorithm addresses the oversight of parameter selection and manages class imbalance without eliminating any legitimate samples. The ensemble framework combines the strengths of convolutional neural networks (CNN) for spatial feature extraction, long short-term memory (LSTM) networks for capturing sequential patterns, and multilayer perceptrons (MLP) for efficient classification. Three ensemble strategies—Weighted average, unweighted average, and unweighted majority voting—are employed to aggregate predictions. Experimental results show that all ensemble methods achieve perfect scores (1.00) in precision, recall, and F1-score for both fraud and non-fraud classes. This study demonstrates the effectiveness of ensemble model with optimized sampling approach for robust and accurate fraud detection
Enhancing sexual education for children with special needs through augmented reality: development and evaluation of the Magical SeDu application
Get PDFThis research focuses on the educational obstacles encountered by children with special needs (CWSN), specifically in sexual education, through developing and evaluating the Magical SeDu application. Using a three-phase instructional design model, the study followed the planning, design, and development phases to create user-centered features that meet diverse learning needs. User acceptance testing (UAT) further confirmed the usability and effectiveness of the app, with a satisfaction rating of 86.04%. These findings underscore the transformative potential of augmented reality (AR) technology in inclusive education, fostering interactive and visually stimulating learning experiences. The study also emphasizes the importance of involving stakeholders in the development process to ensure the app meets the specific needs of its users. Future research should focus on enhancing the app’s features and exploring its integration into broader educational environments to maintain accessibility and continuous improvement. This study contributes to the advancement of inclusive education strategies and highlights the critical role of sex education in increasing self-awareness and protection for children with special needs
AI SWLM: artificial intelligence-based system for wildlife monitoring
Get PDFDetection and recognition of wild animals are essential for animal surveillance, behavior monitoring and species counting. Intrusion of animals and the disaster to be caused can be averted by the timely recognition of intruding animals. An artificial intelligence-based system for wildlife monitoring (AI SWLM) is designed and implemented on the camera trap images. The challenges such as detecting and recognizing animals of different sizes, shape, angles and scale, recognizing the animals of same and different species, detecting them under various illumination conditions, with pose variants and occlusion are addressed by identifying the optimal weights of the deep learning architecture, AI SWLM. Models were trained using Gold Standard Snapshot Serengeti dataset with random weights and the best weights of model were used as initial weights for training the augmented data. This has doubled the performance in terms of mean average precision, which can be interpreted
Cumulative aging effects of five-year intermittent exposure on flexible amorphous solar cells
Get PDFAmorphous silicon (a-Si) is rarely used for large scale photovoltaic energy production, it remains relevant in flexible electronic applications, where mechanical flexibility and lightweight design are prioritized, where exposure to sunlight is typically limited or irregular. This study conducts an experimental analysis of the long-term aging effects on the proprieties of an amorphous solar cells, under five years of intermittent outdoor climate conditions. Unlike conventional aging studies that focus on degradation over time, this research highlights the cumulative effects of environmental exposure, considering the discontinuous nature of exposure cycles and the non-linearity of degradation phenomena because of the abrupt transitions between outdoor exposure phases and indoor laboratory rest periods. The results show that nearly 50% of the panel’s performances is reduced, with the losses observed as follows: a substantial decline in the fill factor from 55.3% to 30%, a decrease in energy conversion efficiency from 11.36% to 5.5%. This accelerated deterioration mainly attributed to harsh environmental transitions caused by intermittent exposure, which amplify aging mechanism compared to continuous exposure. Beyond the experimental findings, the approach presented here, constitutes a meaningful scientific contribution. By introducing a realistic and underexplored aging scenario, it lays the groundwork for a new line of research
Artificial intelligence of things solution for Spirulina cultivation control
Get PDFIn the evolving field of Spirulina cultivation, the integration of the internet of things (IoT) has facilitated the optimization of spirulina growth and significantly enhanced biomass yield in the culture medium. This study outlines a control open-pond system for Spirulina cultivation that employs generative artificial intelligence (AI) and edge computing within an IoT framework. This transformative approach maintains optimal conditions and automates tasks traditionally managed through labor-intensive manual processes. The system is designed to detect, acquire, and monitor basin data via electronic devices, which is then analyzed by a large language model (LLM) to generate precise, context-aware recommendations based on domain-specific knowledge. The final output comprises SMS notifications sent to the farm manager, containing the generated recommendations, which keep them informed and enable timely intervention when necessary. To ensure continued autonomous operation in case of connectivity loss, pre-trained TinyML models were integrated into the Raspberry Pi. These models display alarm signals to alert the farm owner to any irregularities, thereby maintaining system stability and performance. This system has substantially improved the growth rate, biomass yield, and nutrient content of Spirulina. The results highlight the potential of this system to transform Spirulina cultivation by offering an adaptable, autonomous solution
An information retrieval system for Indian legal documents
Get PDFIn this work, a legal document retrieval system is presented that estimates the significance of the user queries to appropriate legal sub-domains and extracts the key documents containing required information quickly. In order to develop such a system, a document repository is prepared comprising the documents and case study reports of different Indian legal matters of last five years. A legal sub-domain classification technique using deep neural network (DNN) model is used to obtain the relevance of the user queries with respective legal sub-domains for quick information retrieval. A query-document relevance (QDR) score-based technique is presented to rank the output documents in relation to the query terms. The presented model is evaluated by performing several experiments under different context and the performance of the presented model is analyzed. The presented model achieves an average precision score of 0.98 and recall score of 0.97 in the experiments performed. The retrieval model is assessed with other retrieval models and the presented model achieves 13% and 12% increase average accuracy with respect to precision scores and recall measures respectively compared to the traditional models showing the strength of the presented model
Enhancing Autonomous GIS with DeepSeek-Coder: an open-source large language model approach
Get PDFLarge language models (LLMs) have paved a way for geographic information system (GIS) that can solve spatial problems with minimal human intervention. However, current commercial LLM-based GIS solutions pose many limitations for researchers, such as proprietary APIs, high operational costs, and internet connectivity requirements, making them inaccessible in resource-constrained environments. To overcome this, this paper introduced the LLM-Geo framework with the DS-GeoAI platform, integrating the DeepSeek-Coder model (the open-source, lightweight version deepseek-coder-1.3b-base) running directly on Google Colab. This approach eliminates API dependence, thus reducing deployment costs, and ensures data independence and sovereignty. Despite having only 1.3 billion parameters, DeepSeek-Coder proved to be highly effective: generating accurate Python code for complex spatial analysis, achieving a success rate comparable to commercial solutions. After an automated debugging step, the system achieved 90% accuracy across three case studies. With its strong error- handling capabilities and intelligent sample data generation, DS-GeoAI proves highly adaptable to real-world challenges. Quantitative results showed a cost reduction of up to 99% compared to API-based solutions, while expanding access to advanced geo-AI technology for organizations with limited resources
Systematic review of artificial intelligence applications in predicting solar photovoltaic power production efficiency
Get PDFThe global energy crisis and climate change demand more accurate and efficient renewable energy forecasting methods. Solar photovoltaic (PV) systems offer abundant clean energy but their efficiency is highly affected by weather variability, requiring advanced predictive models. This systematic review of 69 studies published between 2020 and 2024 evaluates artificial intelligence (AI) and machine learning (ML) applications in PV forecasting, with a focus on hybrid algorithms such as convolutional neural network-long short-term memory (CNN-LSTM). Results demonstrate that hybrid models consistently outperform traditional statistical methods and standalone AI approaches by capturing spatiotemporal patterns more effectively, achieving significant error reductions and improving reliability. A notable gap identified is the limited integration of consumer behavior into forecasting models, despite evidence that incorporating demand-side patterns enhances accuracy. Challenges also remain in data availability, scalability across diverse climates, and computational requirements. This review contributes by synthesizing recent advances and emphasizing consumer integration as an underexplored but critical dimension for future research. The findings provide a foundation for developing more precise, resilient, and scalable PV forecasting models, supporting optimized energy management and accelerating the transition toward sustainable energy systems
Machine learning-based prediction of moisture and oxygen in a large power transformer with online monitoring validation
Get PDFThis study presents a predictive modeling approach for monitoring moisture and dissolved oxygen dynamics in a newly commissioned high-capacity power transformer. Using over 48,000 real-time observations collected across three years via an advanced online monitoring device installed on a 326 MVA generator step-up transformer (GSUT), machine learning models were developed to estimate moisture and oxygen concentrations based on correlated operational parameters. Multiple regression-based algorithms were trained and evaluated using performance metrics including root mean squared error (RMSE), mean absolute error (MAE), and coefficient of determination (R²). Linear regression achieved superior performance with an RMSE values as low as 0.05888 ppm for oxygen and 0.0153 ppm for moisture. The models were further validated using data from a sister transformer, demonstrating generalizability and reliability across similar transformer units. This work contributes a scalable and accurate solution for real-time transformer health assessment, with practical implications for predictive maintenance strategies in power utilities
An integrated FSM-BABER-SROA framework for secure and energy-efficient internet of things networks using blockchain consensus
Get PDFThe rapid expansion of the internet of things (IoT) and wireless sensor networks (WSNs) has intensified the demand for energy-efficient, reliable, and secure data transmission. Traditional clustering and static sleep scheduling approaches often fail to ensure long-term sustainability and tamper-resistant communication. This paper presents BABER-SROAChain, a hybrid optimization and security framework that integrates four core modules: i) Fuzzy similarity matrix (FSM)-based clustering for spatial-energy-aware node grouping, ii) Binary Al-Biruni earth radius (BABER) optimization for intelligent cluster head (CH) selection, iii) ship rescue optimization algorithm (SROA) for adaptive sleep scheduling, and iv) a lightweight blockchain protocol with modified practical byzantine fault tolerance (PBFT) consensus for secure inter-cluster communication. The unified objective function incorporates cluster efficiency, redundancy minimization, latency reduction, and packet delivery ratio maximization. Simulation experiments on large-scale WSNs (100–300 nodes) demonstrate that BABER-SROAChain achieves up to 20% improvement in network lifetime, 18% lower energy consumption, and 15% higher packet delivery ratio compared to state-of-the-art models. Additionally, it minimizes blockchain consensus latency while ensuring high data integrity. The proposed framework offers a scalable, secure, and energy-aware solution suitable for real-time IoT applications, including smart cities, healthcare monitoring, and industrial automation, while addressing the dual challenges of performance optimization and blockchain-based security