Md. Mehedi Hasan

AI Researcher · Behavioral ML · Social Engineering Detection

AI for Social Integrity

I develop interpretable machine learning frameworks to detect social engineering, model behavioral risk, and predict socio-technical system failures. Champion, NASA International Space Apps Champion, Global Nominee & Honorable Mention.

Mehedi Hasan - NASA Space Apps Champion

Team Polaris — NASA Space Apps Challenge 2025 Champions, Global Nommine & Honarable Mention

I was part of Team Polaris, which won the Barisal Division championship. Our team combined expertise in ML, full-stack development, data analysis, and UI/UX design to create a winning solution.

Team Polaris - NASA International Space Apps Challenge 2025
Md. Mehedi Hasan

Research Vision

I work at the intersection of Artificial Intelligence, behavioral modeling, and socio-technical system resilience. My research focuses on:

I aim to develop AI systems that are not only accurate but explainable, fair, and grounded in real human behavior.

Research Projects

SaveFood — Behavioral ML for Food Waste Reduction

Trained XGBoost on IoT time-series to predict spoilage (F1: 0.89). Dashboard visualizes waste reduction impact.

Methods: Time-series forecasting, SHAP interpretability

SafeRoads — Geospatial Risk Prediction

ML model to identify high-risk urban segments using traffic + accident history (Precision: 84%).

Methods: Geospatial ML, ensemble modeling

MeteorShield — Planetary Defense

NASA NEO data platform with real-time 3D visualization. Champion, NASA Space Apps 2025.

Tech: NASA APIs, Three.js, data storytelling

AgroHub — IoT & ML Smart Agriculture

Real-time farm monitoring with ML decision support for smallholder farmers.

Tech: IoT, Python, ML

SciGenie — One-click EDA & AutoML

Accelerates research prototyping with automated EDA and AutoML (Random Forest, XGBoost, Auto-Sklearn).

Tech: Python, Auto-Sklearn, Pandas

Publications

  1. Hasan, M. M., Rakib, R., Molla, M. A., Borhan, R., Based, M. A. — A Socio-Economic Machine Learning Framework for Predicting Programmer Retention. Proceedings of BIM 2025. [Accepted]
  2. Hasan, M. M., Mahin, A. A., Chakraborty, S., Afrose, M., Mia, M. A., Based, M. A. — Behavioral and Demographic Feature Fusion for Developer Attrition Modeling. Proceedings of BIM 2025. [Accepted]
  3. Hasan, M. M., Pallob, M. M. I., Based, M. A. — Digital Transformation for Sustainable Universities: The Strategic Role of Management Information Systems. Unpublished manuscript, 2026.
  4. Molla, M. A., Hasan, M. M., Rakib, R., Borhan, R., Based, M. A. — Explainable Hybrid ML for Rapid Earthquake Event Characterization Using USGS Seismic Metadata. [Under Review]
  5. Mahin, A. A., Hasan, M. M., Mia, M. A., Based, M. A. — EWC-RL++: A Modular and Adaptive Continual Reinforcement Learning Framework for Automation Systems in Structured but Evolving Environments. [Under Review]
  6. Molla, M. A., Hasan, M. M., Rakib, R., Borhan, R., Based, M. A. — Urban--Rural Inequality in Tech Career Intention: A Machine Learning Perspective. [Under Review]
  7. Hasan, M. M., Molla, M. A., Rakib, R., Borhan, R., Based, M. A. — Socio-Economic Determinants of Programming Skill Development in Undergraduate CS Students. [Under Review]
  8. Rakib, R., Molla, M. A., Hasan, M. M., Borhan, R., Based, M. A. — Explainable AI for Understanding Motivation in Choosing Computer Science Careers. [Under Review]
  9. Afrose, M., Mahin, A. A., Hasan, M. M., Mia, M. A., Based, M. A. — Fairness-Aware HIV Risk Prediction for Key Populations in Bangladesh: Bridging Behavioral Surveillance, Explainable AI, and Socio-Technical Accountability. [Under Review]

Skills

Languages
Python
C
Java
SQL
ML & Automation
Scikit-learn
TensorFlow
PyTorch
Selenium
Playwright
RPA
Data & Dev Tools
Pandas
NumPy
Matplotlib
Seaborn
Git
GitHub
VS Code
Colab
Kaggle
Specialized
SHAP
Geospatial
Time-Series
AutoML

Professional Memberships

  • IEEE — Student Member (Valid through Dec 2026)
  • IEEE Computer Society — Student Member (Valid through Dec 2026)
  • BASIS Students' Forum (DIU Chapter) — Executive Member