Article Per Year (5 Year)
p-Index From 2021 - 2026
0.835
P-Index
This Author published in this journals
All Journal Control Systems and Optimization Letters
Shahin Alam
Hubei University of Automotive Technology
Aerospace Engineering Automotive Engineering Computer Science & IT Control & Systems Engineering Electrical & Electronics Engineering Engineering

Published : 5 Documents Claim Missing Document
Claim Missing Document
Check
Articles

Found 5 Documents
Search

 1 
Optimized Hybrid Machine Learning Model for Real-Time Financial Fraud Detection Md Zahidul Islam Sany; Zhang Wubo; Shahin Alam
Control Systems and Optimization Letters Vol 4, No 2 (2026)
Publisher : Peneliti Teknologi Teknik Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.59247/csol.v4i2.332

Abstract

The rapid growth of digital financial transactions has increased the demand for intelligent, scalable, and real-time fraud detection systems capable of identifying fraudulent activities with high accuracy and low latency. This paper proposes an optimized hybrid machine learning framework for real-time financial fraud detection by integrating Random Forest and XGBoost within a weighted soft-voting ensemble classifier. To address the severe class imbalance commonly found in financial datasets, the transaction data were preprocessed using the SMOTE-ENN hybrid resampling technique and Min-Max normalization. Bayesian optimization was employed to tune model hyperparameters and improve generalization while reducing overfitting. The proposed framework was trained and evaluated using stratified data partitioning and 5-fold cross-validation, with performance assessed using Accuracy, Precision, Recall, F1-score, AUC-ROC, and false-negative rate. Experimental results demonstrate that the hybrid ensemble consistently outperforms the individual base classifiers, achieving improved fraud detection capability while maintaining high processing throughput suitable for real-time deployment in a Kafka-based streaming environment. These findings indicate that the proposed framework provides an effective and scalable solution for modern financial fraud detection systems.
Optimized YOLOv11 Architecture for Accurate Multi-Class Vehicle Detection under Real-World Conditions Parag Hossain; Md. Hamim Ferdous; Md. Sabbir Mahmud; Shahin Alam
Control Systems and Optimization Letters Vol 4, No 2 (2026)
Publisher : Peneliti Teknologi Teknik Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.59247/csol.v4i2.316

Abstract

Real-time detection of vehicles is essential for modern transportation systems, traffic surveillance, and autonomous driving technologies. With recent progress in deep learning, object detection models have become more reliable across complex and dynamic environments. This study presents a YOLOv11n (nano variant, 2.59M parameters)-based system designed to accurately detect three key vehicle classes cars, buses, and trucks in real time. Unlike prior YOLO versions, YOLOv11 introduces an improved attention mechanism and anchor-free detection head specifically addressing partially occluded and multi-scale vehicles. A customized dataset (compiled from Open Images v7 and BDD100K, spanning day/night and clear/rain/snow conditions; class distribution: cars 48%, buses 30%, trucks 22%) containing 9,989 training images and 1,998 validation images was used to fine-tune the network. Data augmentation techniques, including Mosaic augmentation, HSV color transformations, and random flips, were applied to enhance model robustness. All experiments were conducted on an RTX 3060 Laptop GPU (6GB VRAM). The trained model achieved strong detection performance, with precision of 0.806, recall of 0.751, mAP50 of 0.830, and mAP50–95 of 0.674 (the 15.6% drop indicates moderate localization errors at stricter IoU thresholds). Inference speed reaches 526 FPS (inference-only) and 263 FPS end-to-end, outperforming YOLOv8n by 3.2% in mAP50. These results highlight YOLOv11n’s ability to balance accuracy and computational efficiency, making it well-suited for real-time applications on resource-limited hardware. The proposed detection framework can support future developments in intelligent traffic management, mobility analytics, and automated road monitoring systems.
Next-Generation 3D Vision and Robotic Integration Technologies for Commercial Vehicle Manufacturing Md Nazmul Islam; Ren Aihua; Muhammad Isyaku; Benfor Nana Boa; Shahin Alam
Control Systems and Optimization Letters Vol 4, No 1 (2026)
Publisher : Peneliti Teknologi Teknik Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.59247/csol.v4i1.284

Abstract

This research project explores the development of 3D machine vision and robot integration technology tailored for commercial vehicle manufacturing and testing. As the industry increasingly demands higher efficiency and quality, the integration of advanced machine vision systems with robotic automation presents a transformative solution. Advanced robotics technologies involve the latest advances in robotics, including integrating sophisticated control systems, sensors, data fusion techniques, and machine learning algorithms. These innovations enable robots to adapt better and interact with their environment, ultimately increasing their accuracy. The main focus of this study is to cover the most common industrial robotic machining processes and to identify how specific advanced technologies can improve their performance. In most of the studied literature, the primary research objective across all operations is to enhance the stiffness of the robotic arm’s structure. This study investigates the implementation of 3D machine vision for precise component inspection, guided assembly, and real-time process adaptation. Key challenges, including high initial costs and complexity of integration, are addressed alongside potential benefits such as improved product quality and reduced operational costs. By leveraging cutting-edge technologies, this research aims to enhance manufacturing processes, ultimately contributing to the advancement of intelligent manufacturing in the commercial vehicle sector. The findings will provide valuable insights for manufacturers seeking to optimize their production capabilities and maintain competitiveness in a rapidly evolving market.
Deep Learning Architectures for Seismic Upgoing–Downgoing Wavefield Separation: A Comparative Benchmark and Cross-Dataset Generalization Study Monirul Islam; Zhong Yu; Shahin Alam
Control Systems and Optimization Letters Vol 4, No 2 (2026)
Publisher : Peneliti Teknologi Teknik Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.59247/csol.v4i2.336

Abstract

Separating upgoing (reflection) energy from downgoing (source-side and multiple) energy is a critical preprocessing step in reflection and vertical seismic profile (VSP) seismology, yet classical frequency–wavenumber, median, and Radon-based filters degrade sharply under lateral velocity variation, topography, and spatial aliasing. This paper reports a systematic, same-dataset comparison of seven wavefield-separation algorithms: a simple convolutional network (CNN), U-Net, bidirectional long short-term memory (BiLSTM), Transformer, ResNet, an MLP with PCA dimensionality reduction, and the classical f–k filter trained and evaluated on 201 synthetic acoustic shot gathers and stress-tested on an independent 240-gather cross-dataset. BiLSTM achieved the best in-distribution performance (correlation = 0.9822, SNR = 14.52 dB) and the smallest relative degradation (41.2%) under domain shift, while U-Net was the strongest convolutional architecture (correlation = 0.7600) and the classical f–k filter performed worst (correlation = 0.3003, SNR = −0.37 dB). All models lost substantial accuracy on the cross-dataset, confirming that domain shift not architectural capacity is the principal barrier to field deployment. The study contributes a reproducible, consistently evaluated benchmark; a rigorous cross-dataset generalization test rarely reported in the literature; and quantitative evidence that recurrent and attention-based sequence models outperform convolutional counterparts for 1-D trace-wise wavefield separation. The findings motivate transfer learning, physics-informed regularization, and larger, more diverse training sets as the next steps toward field-ready deployment.
Real-Time Retail Forecasting and Anomaly Detection Using Hybrid ARIMA and Neural Network Models Khadija Elkattany; Md Mutasim Billah; Shahin Alam
Control Systems and Optimization Letters Vol 4, No 2 (2026)
Publisher : Peneliti Teknologi Teknik Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.59247/csol.v4i2.289

Abstract

This paper presents a hybrid machine learning framework that addresses scalability and accuracy challenges in retail inventory management by integrating real-time demand forecasting with anomaly detection, evaluated using Walmart's historical sales data. Traditional approaches face a trade-off: maintaining individual models for each product category is computationally prohibitive, while generalized models often underperform for dissimilar items, resulting in stockouts or overstocking. To address this, we propose a department-level aggregation strategy that balances specificity and generalization, combined with a hybrid methodology: ARIMA for linear trend and seasonality modeling, cubic spline interpolation to capture nonlinear residual patterns, and neural networks for complex interactions. The framework dynamically adjusts predictions using real-time sales streams and applies residual-based anomaly detection with threshold triggers to identify sudden demand spikes or supply disruptions. Experiments on a filtered Walmart dataset (removing returns, canceled orders, and items with 30 days of historical data; 18 months, 15 departments, aggregated from 100,000+ SKUs) indicate an 18% reduction in mean absolute error (MAE) compared to exponential smoothing baselines (MAE: 235.1 ± 32.8 vs. 310.4 ± 28.5), while spline-enhanced neural networks achieve a 24% improvement over standalone ARIMA (MAE: 235.1 vs. 310.4; p 0.01). The anomaly detection module identifies 92% of simulated irregularities with a 7% false-positive rate and F1-score of 0.89. The proposed framework provides three principal advantages: (1) scalable department-level modeling without per-product customization, reducing training time from 2 hours per product to 12 minutes per department (90% improvement); (2) real-time adaptability to fluctuating demand through 6-hour incremental LSTM updates; and (3) cost-efficient inventory optimization through integrated anomaly alerts, validated in a 6-month pilot across 50 Walmart stores showing 19% stockout reduction and 14% overstocking reduction. This work offers a practical blueprint for retailers to enhance forecasting precision, mitigate supply chain risks, and reduce operational costs in volatile markets.
Co-Authors Benfor Nana Boa Khadija Elkattany Md Mutasim Billah Md Nazmul Islam Md Zahidul Islam Sany Md. Hamim Ferdous Md. Sabbir Mahmud Monirul Islam Muhammad Isyaku Parag Hossain Ren Aihua Zhang Wubo Zhong Yu