Research Statement & Objectives
Seeking PhD opportunities to advance research in Machine Learning and AI Systems
I am actively seeking fully-funded PhD positions in Computer Science with a research focus on machine learning, computer vision, and developing scalable AI systems. My research aims to bridge theoretical machine learning with practical applications, particularly in healthcare, agriculture, and financial systems.
My previous research has demonstrated strong capabilities in developing robust machine learning pipelines for imbalanced datasets, as evidenced by my published IEEE conference papers on stroke prediction and food classification using deep transfer learning. These projects highlight my ability to handle real-world data challenges and develop innovative solutions to complex problems.
My professional background in full-stack development (Django, React, FastAPI) provides me with practical experience in building scalable systems, which complements my research interests by enabling the implementation of production-ready AI solutions. I have extensive expertise in Python, TensorFlow, and various ML frameworks, combined with strong software engineering skills.
Research Goals: I aim to contribute to fundamental AI research by exploring novel approaches in deep transfer learning, handling imbalanced data, and developing explainable AI systems. I am particularly interested in applications that address real-world challenges in healthcare (medical imaging, disease prediction), agriculture (plant disease detection), and financial security (fraud detection).
Why PhD: I am driven by a deep curiosity about artificial intelligence and a desire to contribute to the advancement of the field. My experience in both research (published papers) and practical implementation (full-stack development) positions me well to undertake rigorous PhD research that combines theoretical innovation with practical application.
Research Publications
IEEE Conference Papers and Research Articles
An Innovative Machine Learning Pipeline for Stroke Prediction on Imbalanced Data
This research presents a comprehensive machine learning pipeline addressing class imbalance in stroke prediction datasets. We implemented advanced sampling techniques (SMOTE, ADASYN) and ensemble methods (Random Forest, Gradient Boosting) to achieve significant improvement in prediction accuracy compared to traditional approaches. The pipeline includes feature engineering, model optimization, and thorough evaluation metrics.
IEEE Xplore LinkDry Food Classification using Hybrid Deep Transfer Learning
Proposed a hybrid deep learning approach combining transfer learning with custom CNN architectures for accurate classification of dry food items from images. The model demonstrated state-of-the-art performance on benchmark datasets with 98.2% accuracy, showcasing effective feature extraction and classification capabilities. This work contributes to the field of computer vision for agricultural and food processing applications.
IEEE Xplore LinkAn Efficient Deep CNN-based Approach for Tomato Leaf Disease Detection and Classification
Designed and implemented a deep convolutional neural network for automated detection and classification of tomato leaf diseases with high accuracy. The system provides a practical solution for agricultural monitoring and disease prevention using computer vision techniques. This project demonstrates my ability to apply deep learning to real-world agricultural problems.
Research Projects
Machine Learning and AI Research Implementations
Fraud Transactions Analysis of Mobile Banking using Machine Learning
Undergraduate Thesis Project | Sep 2022 - Apr 2023
Comprehensive analysis of mobile banking fraud detection using various ML algorithms on imbalanced transactional data. Implemented Logistic Regression, Random Forest, Naive Bayes, SVM, Neural Networks, Decision Tree, and KNN with feature engineering and optimization.
Achieved 96.7% precision using Random Forest ensemble methods, implemented advanced sampling techniques (SMOTE, ADASYN) for class imbalance, compared 7 ML algorithms comprehensively, and developed a robust feature engineering pipeline for financial transaction data.
Plant Disease Detection using Deep Learning
Computer Vision Research | 2021-2022
Developed CNN architectures for tomato and mango leaf disease detection using transfer learning with VGG16. Implemented data augmentation and preprocessing techniques for agricultural AI applications with focus on real-world deployment.
Applied transfer learning with VGG16 for mango disease classification achieving 95% accuracy, created custom CNN model with multiple Conv2D layers for tomato diseases, implemented comprehensive data augmentation pipeline, and developed practical solution for agricultural monitoring.
Handwritten Digit Recognition using CNN
Deep Learning Project | 2022
Built CNN model with Conv2D, MaxPooling2D, and Dropout layers for MNIST handwritten digit recognition. Designed architecture with 64 and 32 filter convolutional layers achieving high accuracy on standard benchmark dataset.
This project demonstrates foundational understanding of CNN architectures and their application to image classification problems, serving as building block for more complex computer vision research.
Traffic Sign Recognition System
Computer Vision Project | 2022
Developed CNN model for recognizing 43 traffic sign classes using German Traffic Sign Dataset. Implemented multiple Conv2D layers with 32 and 64 filters, MaxPooling, and Dropout for regularization and improved performance.
Professional & Research Experience
Integrating Software Engineering with AI Research
Backend Developer (AI/ML Focus)
HS Digital Solutions | Dhaka, Bangladesh
- Developing prescription analysis system for NetPharma using Django REST Framework with AI integration for medical data processing
- Built real-time WebSocket chatbot for customer support using AI/NLP models
- Integrating AI models for data analysis and validation in healthcare applications
- Creating scalable backend systems handling high-volume requests for AI-powered platforms
- Collaborating with cross-functional teams to design and implement robust REST APIs
Backend Developer (Remote)
Insidepth Software Solutions | Remote
- Developed AI-powered systems with Django and FastAPI for various applications
- Implemented real-time features using WebSocket and Celery with Redis for AI model inference and data processing
- Built full-stack applications with React.js frontend for ML model deployment and visualization
- Created machine learning APIs for various applications demonstrating research-to-production pipeline
- Collaborated with cross-functional teams to design and implement REST APIs for AI services
Python Trainer & Research Mentor
International Center for Global Skills | Tejgaon, Dhaka
- Conducted Sheikh Russel Digital Lab Python Programming Training 2023 program implemented by Shushilan Limited
- Delivered hands-on coding exercises and Python programming curriculum with ML/AI focus
- Fostered problem-solving skills and guided project development in AI/ML applications
- Mentored 120+ students in Python programming and basic ML concepts
- Improved communication and presentation skills through effective technical content delivery
Technical Implementation Projects
Full-stack applications demonstrating scalable architecture and problem-solving
ChatApp - Real-time Chat SaaS Platform
React JS, FastAPI, WebSocket | 2024
SaaS-based real-time chat solution for businesses featuring WebSocket communication, one-to-one live chat, cross-domain integration, role-based staff management, ticket picking system, customer-staff chat routing, and admin analytics.
FabricTech - Textile Community Platform
React JS, Django, DRF, JWT | 2023
Community-driven platform for the textile industry with secure JWT authentication, multi-role access (admin, author, member), admin dashboard analytics, community/group creation, group posts, courses, events, polls, and author-specific blog publishing.
Multivendor E-commerce Platform
React JS, Django, DRF, JWT | 2023
Full-featured e-commerce platform with multi-payment gateways (PayPal, Stripe), vendor dashboard, admin management, advanced product search, pagination, filtering, rating & review system, and scalable architecture for high traffic.
Restaurant POS System
React JS, Redux, Django, DRF, JWT | 2023
Full-featured restaurant POS system with item management, combo items, variants, ingredient tracking, sales & inventory, customer memberships, multi-branch management, dashboard analytics, barcode scanning, and real-time monitoring.
Education
Academic Background and Research Preparation
Bachelor of Science in Computer Science & Engineering
Bangladesh University of Business and Technology, Dhaka
CGPA: 3.31/4.00 (First Class)
Thesis: "Money Laundering and Fraud Transactions Analysis of Mobile Banking in Bangladesh using Machine Learning"
Location: Mirpur, Dhaka, Bangladesh
Machine Learning, Computer Vision, Artificial Intelligence, Data Mining, Data Structures & Algorithms, Database Systems, Statistics & Probability, Calculus, Linear Algebra, Software Engineering, Computer Networks, Research Methodology, Operating Systems, Computer Architecture
Preprocessed transactional data and engineered features for fraud detection using algorithms like Logistic Regression, Random Forest, Naive Bayes, SVM, Neural Networks, Decision Tree, and KNN. Implemented 7 machine learning algorithms with optimization techniques and achieved 96.7% precision using Random Forest ensemble methods. The research contributed to financial security in mobile banking systems.
Thesis Repository LinkResearch Interests
Areas of expertise and ongoing investigation for PhD research
Medical AI & Healthcare
Developing AI solutions for healthcare applications including medical imaging analysis, disease diagnosis, stroke prediction, and clinical decision support systems with focus on accuracy, reliability, and explainability.
Agricultural AI & Computer Vision
Applying computer vision and deep learning techniques to agricultural problems including plant disease detection, crop classification, yield prediction, and automated monitoring systems for precision agriculture.
Machine Learning Methodologies
Research on advanced ML techniques including imbalanced learning, ensemble methods, transfer learning, and scalable ML systems with applications in finance, security, and real-world problem solving.
Awards & Certifications
Recognitions and professional development
IEEE Conference Publication Certificates
ICICT4SD 2023 & ICCIT 2023 Conferences
Published papers at international IEEE conferences demonstrating research capabilities in deep learning applications and imbalanced data handling. Recognition of contributions to AI research community.
BYLC Leadership Certificate
Bangladesh Youth Leadership Center | 2022
Completed leadership and communication training program demonstrating commitment to professional development and collaborative research environment.
AI Master Class Workshop
30 Days AI Implementation | 2023
Intensive AI model building workshop with practical implementation projects, enhancing hands-on skills in machine learning and deep learning applications.