PhD Student, University of North Texas
I'm a second-year PhD student in the Computer Science and Engineering Department at the University of North Texas. My research interests lie in health informatics, clinical NLP, applied machine learning, and bioinformatics, with a focus on turning unstructured clinical notes into structured signal for patient care. Before this, I spent three and a half years as a software engineer at Dynamic Solution Innovators Limited and Inovace Technologies Limited in Dhaka, Bangladesh.
Pain Severity Classification from Electronic Health Records in Children with Cerebral Palsy
The first systematic benchmark for classifying pain severity in children with cerebral palsy from real-world clinical notes, comparing several natural language processing and language-model approaches to recover clinically meaningful signal from everyday documentation.
Predicting Surgery Likelihood and Outcomes for Children with Cerebral Palsy Using Machine Learning
Uses instrumented gait analysis data and machine learning to predict both the likelihood of surgical procedures and their expected effect on post-operative gait quality in children with cerebral palsy. The combined model pipeline offers clinicians a data-driven reference for comparing surgical options and anticipating outcomes.
Instruction Fine-Tuned Language Model for Pediatric Fall Risk Assessment via Iterative Training Using Generated Clinical Texts
This study introduces the first language-model approach to fall risk assessment in children with cerebral palsy. It turns unstructured clinical notes into clinically meaningful risk signals. The approach relies on validated training methods designed for settings with limited labeled data.
Hybrid Ensemble of Segmentation-Assisted Classification and GBDT for Skin Cancer Detection with Engineered Metadata and Synthetic Lesions from ISIC 2024 Non-Dermoscopic 3D-TBP Images
Hybrid deep learning and GBDT ensemble for skin cancer detection using 3D-TBP images and synthetic lesion augmentation.
An Efficient Android Malware Prediction Using Ensemble Machine Learning Algorithms
Uses an ensemble of machine learning algorithms to predict malicious Android apps from static permission data, reaching 99.5% accuracy with LightGBM.
mVote: A Mobile Voting System to Conduct Election During COVID-19 Pandemic
A mobile voting system designed to let elections continue safely during the COVID-19 pandemic.
Industry Projects
Dynamic Solution Innovators Limited
Backend development for a web-based registrar management system using Java Spring Boot, Thymeleaf, jQuery DataTable, and Jasper Reports.
Dynamic Solution Innovators Limited
Built and published an exam-management app to the Play Store, with MVVM architecture, local storage via SQLite/SharedPreferences, and two product flavors.
Dynamic Solution Innovators Limited
Evaluated ROS, Gazebo, CARLA, and OpenAI Gym for autonomous-vehicle development; simulated lane-following with Canny edge detection and reinforcement learning (Q-learning, DQN, DDPG, TD3, PPO).
Inovace Technologies Limited
Cloud-based attendance monitoring app built with Flutter, Provider state management, SQLite/Hive, and Firebase, including the GP Smart Attendance flavor.
Academic Projects
Undergraduate Capstone Project
Fingerprint-based voter registration and casting, with admin-managed concurrent voting events, scheduling, and result publishing. Built with Java (Android) and Firebase.
Undergraduate Database Project
Relational system modeling entrepreneurs, fishermen, dealers, and cost allotment. PHP/HTML/CSS front end over an Oracle database.
Applying NLP and machine learning to clinical data to improve outcome prediction and care in pediatric healthcare. Working with Shriners Children's Hospital.
CSCE 1040: Computer Science II
CSCE 4115: Formal Languages,
Automata & Computability
