Applied ML
Machine Learning
Risk-scoring frameworks for transaction card fraud, and hybrid models for forecasting — including water-level prediction for the Kaptai Reservoir, where flooding otherwise arrives unannounced.
01Publications
Publications
My research interests are machine learning, deep learning, cyber security and quantum machine learning — applied to problems that matter where I work: scam detection in under-served languages, disease screening from images, and fraud risk scoring.
15 publications · 9 with DOIs · 6 in press · 4 projects · IEEE, Springer, Taylor & Francis
FocusAreas
Research interests
Neural networks
Deep Learning
CNNs and transformers applied to medical and agricultural imaging: pneumonia detection from X-rays, potato tuber disease classification with voting ensembles, and Swin Transformer models for pest damage with explainable AI.
Security
Cyber Security
Phishing, smishing and quishing detection — including Bangla SMS scam detection with BERT and character-level CNNs, for a language nobody else was building tools for.
Quantum
Quantum Machine Learning
Hybrid quantum–classical architectures: quanvolutional networks, residual networks with quantum SVMs, and variational circuits — including where the quantum component genuinely earns its place and where it does not.
Peer reviewedPapers
Peer-reviewed conference papers
Peer-reviewed work in IEEE and Springer proceedings, each with a resolvable DOI. Newest first.
2026
Multi-Layer Email Threat Detection Using N8N Workflow Automation with Real-Time Risk Classification and Automated Response
IEEE International Conference on Quantum Photonics, Artificial Intelligence, and Networking (QPAIN), 2nd edition, CUET, Chattogram · IEEE
2026
Why Classical-Quantum Machine Learning Works Better Than Classical Machine Learning
IEEE International Conference on Quantum Photonics, Artificial Intelligence, and Networking (QPAIN), 2nd edition, CUET, Chattogram · IEEE
2025
Toward Quantum NLP: A Transfer Learning-Based BERT Transformer–QML Framework for Spam Text Semantics Modeling
International Conference on Data Science, AI and Applications (ICDSAIA), Dhaka · Springer
2025
Fine-Tuning Pretrained CNNs for Potato Tuber Disease Classification Using Soft and Hard Voting Ensembles
IEEE International Conference on Robotics, Automation, Artificial-Intelligence and Internet-of-Things (RAAICON) · IEEE
2025
PhishVQC: Optimizing Phishing URL Detection with Correlation Based Feature Selection and Variational Quantum Classifier
International Conference on Intelligent Systems, Advanced Computing and Communication (ISACC) · IEEE
2025
Quantum Machine Learning for Image Classification: A Hybrid Model of Residual Network with Quantum Support Vector Machine
International Conference on Next-Generation Computing, IoT and Machine Learning (NCIM), 2nd edition · IEEE
2025
Quanvolutional Neural Networks for Pneumonia Detection: An Efficient Quantum-Assisted Feature Extraction Paradigm
International Conference on Next-Generation Computing, IoT and Machine Learning (NCIM), 2nd edition · IEEE
2025
A Hybrid Convolutional Neural Network and Random Forest Model for Predicting Water Level Fluctuations in Kaptai Reservoir to Enhance Water Resource Management
International Conference on Electrical, Computer and Communication Engineering (ECCE) · IEEE
2024
Hybrid Machine Learning Model for Detecting Bangla Smishing Text Using BERT and Character-Level CNN
International Conference on Electrical and Computer Engineering (ICECE) · IEEE
Complete record: Google Scholar · ORCID · ResearchGate
In pressAccepted
Accepted and in press
Accepted for publication, awaiting proceedings.
2026
AgriLM: An Intelligent Agricultural IoT Platform for Climate Change Adaptation Using Large Language Model
International Conference on Sustainable Agriculture and Sustainable Development (ICSASD) · Taylor & Francis · in press
2026
Precision Agriculture for Jujube: Swin Transformer Based Early Identification of Insect Pest Damage Using Explainable AI
International Conference on Sustainable Agriculture and Sustainable Development (ICSASD) · Taylor & Francis · in press
2026
Quantum Machine Learning as a Transformative Frontier for Sustainable Smart Agriculture: Opportunities, Challenges and Future Directions
International Conference on Sustainable Agriculture and Sustainable Development (ICSASD) · Taylor & Francis · in press
2026
Fine-Tuned DenseNet121 with BORUTA for Potato Tuber Disease Classification confirm authors
SPICSCON 2026, BAUET, 13–14 August 2026 · IEEE · accepted, camera-ready
2025
CLocMLGuard: A Novel Risk Scoring Framework for Transaction Card Fraud Mitigation and Detection with Machine Learning
International Conference on Computing, Communication and Networking Technologies (ICCCNT), 16th edition, Indore · IEEE · in press
BuiltProjects
Things I've built
Deployed applications and machine learning prototypes — most small enough for one person to finish, and useful enough for someone to actually run. Everything below has a live demo or public source.
Machine learning · Health
Breast Cancer Diagnosis
A diagnostic classifier trained on the 1995 Breast Cancer Wisconsin (Diagnostic) dataset, running entirely in the browser: Python and scikit-learn are compiled to WebAssembly and loaded client-side, so there is no server and nothing to install. First load takes 15–40 seconds while the runtime downloads; after that predictions are instant.
Built as a research and teaching prototype. It is explicitly not a medical device, and the interface says so — outputs must not be used to make or withhold a diagnosis.
Pythonscikit-learnFastAPIGradioDockerWebAssemblyHugging Face Spaces
Web application
Abdullah Tech Shop
A shop management system built in Flask for my uncle's business — stock, sales and reporting for a shop that had been run on paper. I wrote up the build and the deployment as a blog post.
FlaskPythonRender
Web application
M.R.S. Easy Pay
A payment portal for M.R.S Computer. Customers open it to copy the shop's bank account details cleanly rather than transcribing them from a message, which is where transfer errors usually come from. Mobile-first, and deliberately single-purpose.
ReactTypeScriptViteTailwindshadcn/ui
Machine learning · Energy
Solar Cell Efficiency Prediction
A notebook predicting photovoltaic cell efficiency from cell characterisation data — a regression problem on measured device parameters rather than image data.
JupyterPython
AlsoClassroom
Teaching repositories
Database CRUD exercises written for my WACE students, kept public so they can clone them and work through the tasks at home.
StackSkills
Tools and methods
What I build with day to day, grouped by what it is for. Deep learning and deployment for the model work, quantum frameworks for the hybrid circuits, and the ordinary web and data tooling that gets a model in front of someone.
Languages & ML
Quantum
Deployment
Web & data
Environment
Writing
ServiceReview
Peer review
Article reviewer for the Asia Pacific Journal of Energy and Environment since 2025, providing clear, unbiased, and timely feedback on submitted manuscripts and advising editors on acceptance or rejection.
AimVision
Where I want this to go
My ultimate goal is to research in machine learning, focusing on developing interactive solutions and leveraging technology to solve common challenges, particularly those faced by people in developing countries.
Looking forward, I aspire to become a successful machine learning researcher, dedicated to advancing research initiatives that contribute to the development of Bangladesh. Additionally, I aim to represent Bangladesh on the global stage, fostering collaborations with developed nations and supporting the growth of young Bangladeshi researchers in the field of machine learning.