Article Per Year (5 Year)
p-Index From 2020 - 2025
1.547
P-Index
This Author published in this journals
All Journal Journal of Applied Data Sciences International Journal of Engineering, Science and Information Technology
Batumalay, M.
Unknown Affiliation
Astronomy Biochemistry, Genetics & Molecular Biology Chemical Engineering, Chemistry & Bioengineering Chemistry Civil Engineering, Building, Construction & Architecture Computer Science & IT Control & Systems Engineering Decision Sciences, Operations Research & Management Earth & Planetary Sciences Education Electrical & Electronics Engineering Energy Engineering Industrial & Manufacturing Engineering Library & Information Science Materials Science & Nanotechnology Mathematics Mechanical Engineering Physics Social Sciences Transportation

Published : 10 Documents Claim Missing Document
Claim Missing Document
Check
Articles

Found 10 Documents
Search

 1 
Combined Fire Fly – Support Vector Machine Digital Radiography Classification (FF-SVM-DRC) Model for Inferior Alveolar Nerve Injury (IANI) Identification Manikandaprabhu, P.; Thirumoorthi, C.; Batumalay, M.; Xu, Zhengrui
Journal of Applied Data Sciences Vol 5, No 3: SEPTEMBER 2024
Publisher : Bright Publisher

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47738/jads.v5i3.356

Abstract

Inferior Alveolar Nerve Injury (IANI) is a severe complication in oral surgery that can significantly affect a patient's quality of life. Accurate diagnosis is crucial for effective management, and digital radiography has become an essential tool in this regard. This study proposes a novel feature selection-based classification algorithm to enhance the diagnostic precision of digital radiographs (DRs) for IANI detection. The objective is to improve classification accuracy by selecting the most relevant features using a Firefly algorithm-based method. Our approach identifies optimal features that preserve critical information from the dataset, enabling more accurate predictions by machine learning models. The proposed method was tested using a dataset of 140 DRs and achieved a classification accuracy of 97.4%, with a sensitivity of 80.9% and a specificity of 94.8%. These results demonstrate that the Firefly algorithm-based feature selection significantly outperforms traditional methods in diagnosing IANI. The novelty of this research lies in its integration of advanced feature selection techniques with support vector machines, offering a robust tool for improving diagnostic accuracy in dental imaging. This work contributes to enhanced clinical decision-making and could be valuable for broader applications in healthcare systems.
Deep Wiener Deconvolution Denoising Sparse Autoencoder Model for Pre-processing High-resolution Satellite Images Kiruthika, S.; Priscilla, G. Maria; Vijendran, Anna Saro; Batumalay, M.; Xu, Zhengrui
Journal of Applied Data Sciences Vol 5, No 3: SEPTEMBER 2024
Publisher : Bright Publisher

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47738/jads.v5i3.357

Abstract

The detection of geospatial objects in surveillance applications faces significant challenges due to the misclassification of object boundaries in noisy and blurry satellite images, which complicates the detection model's computational complexity, uncertainty, and bias. To address these issues and improve object detection accuracy, this paper introduces the Deep Wiener Deconvolution Denoising Sparse Autoencoder (DWDDSAE) model, a novel hybrid approach that integrates deep learning with Wiener deconvolution and Denoising Sparse Autoencoder (DSAE) techniques. The DWDDSAE model enhances image quality by extracting deep features and mitigating adversarial noise, ultimately leading to improved detection outcomes. Evaluations conducted on the NWPU VHR-10 and DOTA datasets demonstrate the effectiveness of the DWDDSAE model, achieving notable performance metrics: 96.32% accuracy, 86.88 edge similarity, 75.47 BRISQUE, 28.05 IQI, 38.08 PSNR (dB), 0.883 SSIM, 98.25 MSE, and 0.099 RMSE. The proposed model outperforms existing methods, offering superior noise and blur removal capabilities and contributing to Sustainable Development Goals (SDGs) such as SDG 9 (Industry, Innovation, and Infrastructure), SDG 11 (Sustainable Cities and Communities), and SDG 13 (Climate Action). This research highlights the model's potential for inclusive innovation in object detection applications, showcasing its contributions and novel approach to addressing existing limitations.
IoT based Intrusion Detection for Edge Devices using Augmented System Nagarajan, R.; Batumalay, M.; Xu, Zhengrui
Journal of Applied Data Sciences Vol 5, No 3: SEPTEMBER 2024
Publisher : Bright Publisher

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47738/jads.v5i3.358

Abstract

The Edge Computing (EC) paradigm is gaining popularity among users due to its inherent characteristics and expeditious delivery approach. Users may get information from the network's edge thanks to this feature of network architecture. The security of this edge network design, however, is a major issue. Through the Internet and in a shared setting, users can access all EC services. Intrusion detection is a method of network security that searches for threats. It is ineffective to monitor real-time network data, and current detection techniques are unable to identify known dangers. To address this problem, a technique known as augmentation oversampling is proposed, which incorporates the minority classes in the dataset. Our Sort-Augment-Combine (SAC) approach divides the dataset into subsets of the class labels, from which synthetic data is generated for each group. The developed synthetic data was then used to oversample the minority classes. After the oversampling process was complete, the distinct classes were combined to provide improved training data for model fitting. When compared to the original dataset, the models trained using the enhanced datasets perform better in terms of accuracy, recall (sensitivity), and true positives (specificity). SAC fared best in a UNSW-NB15 dataset when compared to the Synthetic Minority Oversampling Technique (SMOTE) and Generative Adversarial Network-Data Augmentation (GAN-DA). Additionally, SAC points to improvements in general sensitivity, specificity, and accuracy. SMOTE, datasets with ROSE enhancements, and Random Over-Sampling Examples for process innovation.
Impact of FACTS Devices on Reactive Power Optimization in Hybrid Renewable-Grid Networks Rajasree, R.; Lakshmi, D.; Batumalay, M.
Journal of Applied Data Sciences Vol 6, No 3: September 2025
Publisher : Bright Publisher

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47738/jads.v6i3.743

Abstract

Renewable energy integration with conventional electric power networks creates power-quality and stability difficulties because of their inherent volatility. The reliability improvement of hybrid renewable-grid systems depends heavily on reactive power optimization for achieving voltage control as well as loss reduction. The research explores the application of Flexible AC Transmission System (FACTS) devices with special emphasis on Distribution Static Compensator (DSTATCOM) devices for distributing reactive power compensation at the distribution level. The optimization process utilizes Particle Swarm Optimization (PSO) because it demonstrates both quick convergence and strong abilities for global search within nonlinear systems. The PSO algorithm functions to determine the perfect settings of the DSTATCOM device that enables voltage regulation within safety bounds and improves power factor performance. The hybrid system connects PV array components with wind turbines for power management together with the main grid while dealing with fluctuating load requirements. Under optimized conditions simulation output shows that DSTATCOM reduces reactive power requirements in substantial amounts. DSTATCOM's implementation enables the system to achieve better voltage security together with diminished power losses and superior load power factor levels. Detailed research shows that DSTATCOM proves efficient while being attached to the main grid for real-time compensation operations. The PSO system enables it to function efficiently throughout changing conditions of power generation and load requirements. Smart grid efficiency along with resilience advances because of the combined operation of FACTS devices and swarm intelligence methods. Through its proposed method the system ensures lasting grid sustainability and manages renewable resources intermittency effectively for process innovation.
Data-Driven Optimization of UPQC Performance for Solar PV Systems in Weak Grids Using Simulation and Predictive Modeling Rajasree, R.; Lakshmi, D.; Batumalay, M.
Journal of Applied Data Sciences Vol 6, No 3: September 2025
Publisher : Bright Publisher

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47738/jads.v6i3.742

Abstract

The integration of solar photovoltaic (PV) systems into weak power grids presents significant challenges due to low short circuit ratios (SCR), resulting in voltage instability, high harmonic distortion, and diminished fault tolerance. This study proposes a data-driven framework to enhance grid stability and power quality by employing a Unified Power Quality Conditioner (UPQC) integrated with Proportional-Integral (PI) controllers. A comprehensive simulation model was developed using MATLAB/Simulink and validated through hardware-in-the-loop (HIL) experiments. Key electrical performance metrics—such as voltage profiles, total harmonic distortion (THD), and reactive power—were collected and analyzed. To enhance system insight, the dataset was further processed using statistical analysis and predictive modeling techniques to evaluate control response under varying solar irradiance and load conditions. The results demonstrate that the UPQC system maintains stable voltage, reduces THD to within IEEE-519 standards, and improves power factor to 0.98. This research highlights the potential of combining power electronics control with data-centric evaluation to ensure reliable renewable energy integration in weak grid environments. The proposed system contributes toward developing intelligent grid-support solutions for sustainable energy transitions and process innovation.
A Hybrid CNN-Transformer Model with Quantum-Inspired Fourier Transform for Accurate Skin Disease Classification S, Aasha Nandhini; Manoj, R. Karthick; Batumalay, M.
Journal of Applied Data Sciences Vol 6, No 3: September 2025
Publisher : Bright Publisher

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47738/jads.v6i3.782

Abstract

Skin disease classification is a complex task that requires robust feature extraction, efficient classification, and interpretability. Artificial intelligence-based technologies offer effective solutions for developing a framework for skin disease classification while ensuring explainability for healthcare professionals. This study proposes a novel Hybrid Transformer model comprising of Convolutional Neural Network (CNN) architecture infused with a Quantum-Inspired Fourier Transform (QIFT) to enhance classification accuracy. QIFT is incorporated to emphasize frequency-domain information alongside the spatial features captured by CNNs, potentially improving feature representation and model generalization. For demonstration, a dataset containing four different classes of dermatological images is used. Data augmentation techniques and adaptive learning rate scheduling are employed to optimize the dataset. A weighted cross-entropy loss function is used to address class imbalances in the dataset. In this research, explainability is implemented using a standard attribution technique like Integrated Gradients providing insights into model decision-making, and enhancing trust in medical applications. Performance evaluation involves validating the proposed framework using metrics such as confusion matrix analysis, classification reports, and training-validation curves. Experimental results demonstrate a high classification accuracy of 92.5% across skin disease categories. The findings indicate that integrating QIFT and CNN-based feature extraction with transformer-driven attention mechanisms enhances skin disease classification performance while ensuring interpretability as process innovation.
Design of Ethical AI Frameworks for Sustainable and Adaptive Energy Management Systems Humadi, Mustafa; Abbas, Haider Hadi; Hilou, Hassan Waryoush; Najm, Nahlah. M. A. D.; Ali, Ammar Abdulkhaleq; Batumalay, M.
International Journal of Engineering, Science and Information Technology Vol 5, No 1 (2025)
Publisher : Malikussaleh University, Aceh, Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.52088/ijesty.v5i1.1288

Abstract

The integration of Artificial Intelligence (AI) in Energy Management Systems changed completely how sustainable infrastructure operates?and is guarded. But the growing independence of AI decision-making presents some serious ethical questions about?fairness, transparency, and accountability. The article introduces a new framework with Ethical AI for Sustainable and Adaptive Energy Management Systems (EAI-SEM) that is designed to combine functional (re)configuration for operational control and ethical governance in centralized: smart buildings and?decentralized: nano-grid settings. The approach incorporates deep reinforcement learning for adaptive control, federated learning for privacy-preserving model updates, and an?integrated Ethics Verification Module for a dynamic assessment of privacy-conformance levels. In experimental simulations over 30-day operation of the smart building and 10-rounds of federated training of the nano-grid, unjust fairness deviation and explainability of the system experienced enhancements, which also indicated?the reduction of carbon dioxide emissions. The?study demonstrated that ethical protocols can be included without impacting on computational efficiency and system responsiveness. Additionally, the federated structure facilitated decentralized ethical responsibility across different actors and thus allowed for the scalable?implementation. The authors verify the possibility of integrating ethics into the computational core of?intelligent energy systems, near from auditing static policies, towards dynamic ethical choices. In the future the process innovation work could be applied to deployments in other infrastructure systems like water?systems and mobility systems, and it provides a reproducible model for the embedding of normative reasoning into AI for infrastructure.
Neuromorphic Hardware Design for Energy-Aware Artificial Intelligence Computation Aljanabi, Yaser Issam Hamodi; Hussain, Salah Yehia; Salim, Darin Shafiq; Al-Doori, Vian S.; Brieg, Jassim Mohamed; Batumalay, M.
International Journal of Engineering, Science and Information Technology Vol 5, No 1 (2025)
Publisher : Malikussaleh University, Aceh, Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.52088/ijesty.v5i1.1279

Abstract

Rapid growth of the energy-efficient artificial intelligence (AI) systems has attracted substantial interest in neuromorphic computing that emulates organization and actions of a biological neural?system to support low-power, event-driven information processing. In this work, we propose a neuromorphic hardware architecture for energy-efficient AI computing that utilizes spiking neural networks and monolithic?vertical integration to improve the performance of a variety of vision tasks. The architecture is tested against three benchmark datasets— MNIST, N-MNIST, and DVS128,?representing static, spiking and dynamic input modalities, respectively. The performance metrics, such as energy efficiency, inference latency,?throughput, classification accuracy, and unified Energy Efficiency Index (EEI) are compared to characterize the generalization power of the system in different processing environments. Experimental results show that the proposed chip provides a sharply lower energy per inference with a competitively performing accuracy over conventional AI?accelerators, including GPU-based and microcontroller platforms. Additionally, the hardware achieves sub-2 ms inference latency and high throughput, indicating suitability for real-time, embedded AI applications. Comparative analysis with existing neuromorphic platforms highlights the advantage of architectural co-design in balancing energy and performance constraints. While the absence of on-chip learning presents a limitation, the system offers a scalable foundation for edge AI systems requiring efficient, continuous inference. Future directions include integrating adaptive learning mechanisms and extending evaluation to broader AI domains as a process innovation.
Federated Learning Architectures for Privacy-Preserving Smart Grid Data Processing Abdulkareem, Sarah Ali; M. Kallow, Sabah; Bako, Imad Matti; Abdullah, Salima Baji; T.Y. Alfalahi, Saad; Batumalay, M.
International Journal of Engineering, Science and Information Technology Vol 5, No 3 (2025)
Publisher : Malikussaleh University, Aceh, Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.52088/ijesty.v5i3.1423

Abstract

The use of smart data in smart grid infrastructure has lately become essential for efficient power distribution, instantaneous?decision-making and overall system protection. Nonetheless, the application of centralized machine-learned models is impeded by?privacy issues, nonhomogeneous distributed data sources, and communication constraints. In this paper, we propose a federated learning framework to handle these challenges and support decentralized, privacy-preserving?model training across a wide range of smart grid components such as residential meters, substations, and electric vehicle charging stations. The proposed method develops a multi-staged framework, which includes adaptive differential privacy, gradient compression, and topology-aware aggregation to improve?the model's performance in the meanwhile of data privacy. The robustness of the system is demonstrated by energy profiling, cross-domain generalization test and temporal?stability analysis. Findings indicate the model has good prediction performance across different grid setups and customer profiles and that energy use and privacy?noise are within acceptable limits for operational use. Furthermore, the architecture shows?strong generalization to unseen domains, and robust performance through many federated training rounds. By considering?computational efficiency, privacy limitations and topological heterogeneity, this work provides a scalable and secure real-time energy intelligence approach. Results suggest that federated?learning with adaptations to the smart grid is a promising approach for robust privacy-preserving analytics applied to critical infrastructures. This work will support energy efficiency in the future which will be a process innovation. 
IoT-Enabled Supervised Learning-Based Prediction Model for Smart Instrumentation Controllers in Signal Conditioning Systems Prakash, S.; Kalaiselvi, B; Batumalay, M.
Journal of Applied Data Sciences Vol 6, No 4: December 2025
Publisher : Bright Publisher

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47738/jads.v6i4.803

Abstract

This study proposes an intelligent Machine Learning (ML)-based smart controller for industrial flow process systems to enhance accuracy, adaptability, and robustness compared to conventional Proportional–Integral–Derivative (PID) controllers. The main idea is to replace reactive PID tuning with a proactive data-driven control strategy capable of predicting deviations and adjusting process parameters in real time. The objective is to develop and evaluate supervised learning models that can replicate and improve PID performance using real-time operational data collected from a flow process station. The proposed system integrates Internet of Things (IoT) sensors and edge computing to continuously acquire and process flow rate, pressure, and valve position data for model training and testing within the WEKA platform. Four classifiers—Linear Regression, Multilayer Perceptron (MLP), Sequential Minimal Optimization Regression (SMOreg), and M5P model tree—were compared using Mean Squared Error (MSE), Root Mean Squared Error (RMSE), Relative Absolute Error (RAE), and model-building time as key evaluation metrics. Experimental results demonstrated that the M5P pruned tree model achieved the best overall performance with an MSE of 0.0024, RMSE of 0.0577, and model-building time of only 0.03 seconds, outperforming Linear Regression (RMSE = 0.0028), MLP (RMSE = 0.026), and SMOreg (RMSE = 0.0279). The findings show that the M5P-based controller closely replicates PID behavior while offering superior predictive accuracy, faster computation, and self-adaptive learning capabilities. The novelty of this research lies in demonstrating that an IoT-enabled, data-driven smart controller can achieve real-time predictive control without requiring explicit mathematical models, thereby simplifying tuning complexities and paving the way for autonomous, scalable, and intelligent control systems in Industry 4.0 environments.
Co-Authors Abbas, Haider Hadi Abdulkareem, Sarah Ali Abdullah, Salima Baji Al-Doori, Vian S. Ali, Ammar Abdulkhaleq Aljanabi, Yaser Issam Hamodi Bako, Imad Matti Brieg, Jassim Mohamed Hilou, Hassan Waryoush Humadi, Mustafa Hussain, Salah Yehia Kalaiselvi, B Kiruthika, S. Lakshmi, D. M. Kallow, Sabah Manikandaprabhu, P. Manoj, R. Karthick Nagarajan, R. Najm, Nahlah. M. A. D. Prakash, S. Priscilla, G. Maria Rajasree, R. S, Aasha Nandhini Salim, Darin Shafiq T.Y. Alfalahi, Saad Thirumoorthi, C. Vijendran, Anna Saro Xu, Zhengrui