Abdur Rahman Fahad

I am a graduate student in Computer Science at Missouri State University, pursuing my M.S. under the supervision of Dr. Razib Iqbal. My research interests include Cyber-Physical Systems (CPS), Internet of Things (IoT), smart environments, computer vision, and deep learning applications across diverse domains.

I have over two years of industry experience as a Software Engineer (ML/AI) working on large-scale biometric recognition systems. I completed my undergraduate degree in Computer Science and Engineering from Bangladesh University of Engineering and Technology (BUET).

Education

Missouri State University

Master of Science

Computer Science

CGPA: 4.00 / 4.00

Relevant Coursework:
Information Retrieval, Explainable AI, Deep Learning, Computer Vision, Software Test/Quality Assurance.

August 2024 - May 2026 (Expected)

Bangladesh University of Engineering and Technology (BUET)

Bachelor of Science

Computer Science and Engineering

Relevant Coursework:
Human Computer Interaction, Microprocessors Microcontrollers and Embedded System, High Performance Database System, Pattern Recognition, Computer Networking, Computer Security, Computer Graphics, Software Engineering, Theory of Computation, Digital Logic Design, Operating System, Computer Architecture, Compiler, etc.

Feb 2017 - May 2022

Publications

A Metamorphic Testing Framework for Verification and Validation of Unsupervised Sensor Grouping in Smart Spaces via Spectral Clustering

Fahim Ahmed Irfan, Md Asif tanvir, Md Abdur Rahman Fahad, Razib Iqbal

In this paper, we propose a meta- morphic testing approach for verifying a novel sensor grouping technique in smart spaces. The technique employs a spectral clustering algorithm with graph-based feature representations derived from time-series data. Our approach defines ten metamorphic relations encompassing both verification and validation perspectives.

STVR Journal
(Under Review)

Towards Human-Centric Smart Homes: Modeling Sensor-Actuator Interactions with Deep Learning

Md Abdur Rahman Fahad, Razib Iqbal

In this paper, we propose a novel data-driven approach for learning sensor-actuator activation and deactivation patterns in smart home environments, independent of specific human activities. Our method extracts information on sensor-actuator interactions, which is leveraged to generate adaptive operational policies for automation.

ACM HUMANSYS 2025

Design and Implementation of a Scalable Clinical Data Warehouse for Resource-Constrained Healthcare Systems

Shovito Barua Soumma, Fahim Shahriar, Umme Niraj Mahi, Md Hasin Abrar, Md Abdur Rahman Fahad, Abu Sayed Md. Latiful Hoque

This study proposes a scalable, privacy-preserving clinical data warehouse, designed for heterogeneous EHR integration in resource-constrained settings and tested with 1.16 million clinical records. The system supports automated multi-source ingestion, phonetic patient matching, and disease-specific analytics through modular data marts.

IEEE EMBC 2025

Research

Automating Human Centric Operational Policy Generation For Sensor-Actuator Network in Smart Environments

Missouri State University | Advisor: Dr. Razib Iqbal

My graduate thesis focuses on developing adaptive smart home automation frameworks by modeling sensor-actuator interactions beyond rule-based automation. Currently, I am working on achieving higher prediction accuracy with reduced computational overhead, making the models practical for resource-constrained IoT and edge devices.

August 2024 - present

Computational Learning Systems Lab

Collaboration with Dr. Tayo Obafemi-Ajayi

Worked on modeling and processing mTBI (FITBIR) dataset to identify patterns and cluster patient profiles.
Worked on developing a Automatic Clustering framework for algorithm selection, genetic algorithm based hyperparameter tuning, and clustering evaluation using multiple CVIs.
Developing interpretable ECG image classification models with SHAP-based explainability as part of the ROSE 2025 Summer Project.

January 2025 - present

Biometric Recognition Systems Research

R&D Department, TigerIT Bangladesh Ltd

I contributed to the development and optimization of core solutions to the Company's biometric identity management systems, including fingerprint, face, and iris recognition solutions. Also worked in Hip Dysplasia detection from Ultrasound images as a part of external collaboration.

June 2022 - July 2024

Performance Comparison Between PostgreSQL and Hbase for Data Warehouse

Advisor: Dr. Abu Sayed Md. Latiful Hoque, Professor, CSE, BUET

As part of undergraduate thesis we performed in depth analysis on performance of relational (PostgreSQL) and non-relational (HBase) database systems for large-scale data warehouse applications in the National Clinical Data Warehouse (NCDW). We conducted analytics and benchmarking to measure query latency, scalability, and storage efficiency across multi-dimensional datasets.

June 2021 - April 2022

Experience

Graduate Research Assistant

Missouri State University

In MuSyC Lab I am working towards policy generation for user behavior based automation of Smart Appliances.

Served as Problem Setter and Judge for 2025 CODERS Olympiad.
Collaborated with Dr. Keiichi Yoshimatsu on Graphormer-IR-based prediction of molecular permeability and infrared spectra from SMILES-encoded structures.
Helped with the Grant Renewal Proposal of NSF-funded ROSE Project focusing on latest NSF priorities.

August 2024 - Present

Graduate Mentor - ROSE Research Experience for Teachers (RET) Program

NSF ROSE Grant

Mentored K-12 teachers in coding and applied machine learning as part of the NSF-funded ROSE RET program. Taught Python programming, introduced AI and ML fundamentals, and guided hands-on sessions on building models using KNIME and AutoML frameworks. Supervised a research project on ECG image classification using the PTB-XL dataset and interpreting predictions with SHAP-based explainability.

Summer 2025

Software Engineer (ML/AI)

TigerIT Bangladesh Ltd

Worked in the Research and Development division of the country's leading biometrics company, focusing on advancing large scale fingerprint, face, and iris recognition systems ranked among top NIST and FVC benchmarks.

Worked on deep learning models for partial fingerprint reconstruction, including ridge orientation and template generation from incomplete or noisy images using self-supervised learning.
Developed lightweight, high-accuracy face embedding models (EER = 0.06028%) optimized for edge device deployment, built SDK for Android.
Developed semantic segmentation models for iris, pupil, and eyelash detection.

June 2022 - July 2024

Projects

Graduate Projects

Scene Graph Generation for Object Localization in Smart Homes from User Commands

Course: CSC737 - Deep Learning

Developed a scene graph generation system that interprets user commands and locates corresponding objects from a camera feed. Utilized DistilBERT for context extraction from natural language queries, Detectron2 for object detection and localization, and an iterative transformer-based scene graph generator for relational reasoning between entities. Trained and evaluated the model using the Visual Genome dataset for context-aware object localization.

Generating Pareto-Optimal Counterfactuals through Evolutionary Multi-Objective Optimization

Course: CSC790 - Explainable AI

In this project we enriched DiCEML framework by integrating PyMOO, a powerful library for multi-objective optimization, to generate higher quality counterfactual explanations. By implementing algorithms like NSGA-II as a new sampling strategy, better balance was achieved in competing objectives such as proximity, diversity, and sparsity in the explanations it produced.

Movie Review Sentiment Classifier using Naive Bayes and Feature Selection Metrics

Course: CSC790 - Information Retrieval

For this project, we built a movie review sentiment classifier leveraging Naive Bayes and information retrieval based feature selection techniques. We used Mutual Information, Chi-Square, and Collection Frequency to identify high-impact terms, enhancing model precision and interpretability.

Undergraduate Projects

RideSharingApp-TakeME

Course: CSE307,308 - Software Engineering

Developed a full-stack ride-sharing app consisting of three Android apps - for users, drivers, and vehicle owners. Implemented RESTful API communication using Volley, JWT based authentication, and Mapbox for live location tracking and vehicle management. Integrated real-time updates, ride lifecycle management and dashboard analytics for each role.

3D Rendering Framework Using Ray Tracing in OpenGL

Course: CSE 409,410 - Computer Graphics

Built a C++ ray tracer capable of rendering realistic 3D scenes with multi-level reflections. Applied Phong lighting to model ambient, diffuse, and specular illumination. Designed camera controls and geometry handling for dynamic scene visualization.

Rice Crop Disease Detection

Course: CSE471,472 - Machine Learning

Built a model using InceptionV3 transfer learning to classify rice crop diseases and pests from field images collected in Bangladesh rice fields. The dataset included categories like BLB, Brown Spot, Neck Blast, Sheath Blight Rot, Stem Borer, along with healthy plant samples.

Minimal C Compiler

Course: CSE309,310 - Compiler

A custom compiler for a subset of the C language using Flex for lexical analysis and Bison for syntax and semantic parsing. Implemented multi-stage compilation to generate equivalent 8086 assembly code, including data declarations, control flow translation, and arithmetic operations. Integrated error handling for lexical, syntax, and semantic violations, outputting detailed logs for debugging.

Smart Traffic Signal Control System

Course: CSE315,316 - Microprocessors, Microcontrollers, and Embedded Systems

Adaptive traffic signal control system using ATmega32 microcontrollers and HC-SR04 ultrasonic sensors to detect real-time vehicle congestion. Integrated multi-microcontroller communication, timer-based jam detection, and automated signal scheduling.

DoS attack to the DNS server

Course: CSE 405,406 - Computer Security

Denial-of-Service (DoS) attack simulation on a local DNS server configured with Bind9. Implemented an IP spoofing mechanism to generate randomized UDP-based DNS queries, exhausting server's cache resources. Used Wireshark to analyze packet behavior, validate spoofed headers, and demonstrate complete service denial.

Please feel free to visit my github profile for more projects.

Skills

Programming Languages & Tools

Programming: C/C++, Python, Java, Bash, SQL, HTML/CSS, Assembly
Miscellaneous: Pytorch, NumPy, Pandas, OpenCV, Docker, Angular, Oracle, PostgreSQL, MongoDB, Scikit-learn, Android Development, Latex
Soft Skills: Problem Solving, Teamwork, Documentation, Task Organization

Achievements

Graduate Thesis Funding Award, Missouri State University, 2025
Poster Presentation, 32^nd Interdisciplinary Forum, 2025
7th in Buet Programming Competition, CSE FEST, 2018
University Merit Scholarship, BUET, 2017
Government Scholarship - Higher Secondary Certificate Exam, 2016 ( Notre Dame College, Dhaka )
Government Scholarship - Secondary School Certificate Exam, 2014 ( Rajuk Uttara Model College, Dhaka )
Government Scholarship - Primary Scholarship Examination, 2008 ( Uttara High School & College )
Solved numerous problems on Codeforces, Toph , HackerRank, CodeChef, and LeetCode.

Contact

Room 213A (MuSyC Lab), Cheek Hall
901 S National Ave
Springfield, MO 65897

Email: fahad110490@gmail.com, mf8494s@missouristate.edu