Talk to Me

Shahan Ahmed

Computational Social Science · Natural Language Processing · Healthcare Analytics

I am a researcher in computational social science, applying natural language processing and adversarial machine learning to questions of healthcare equity and public health analytics.


Research Interests

  • Computational Social Science
  • Natural Language Processing & Text Classification
  • Adversarial Machine Learning & Robustness
  • Healthcare & Public Health Analytics
  • Data Infrastructure for Social Research

Research

Social Engineering & Adversarial Obfuscation in BEC Attacks

2026 · 95.4% detection accuracy

A character-level classifier flagging Business Email Compromise attempts obfuscated with Unicode homoglyphs and zero-width characters.

[View results]

Robustness of Phishing Detection Under Adversarial Unicode Obfuscation

2026 · 99.8% accuracy

An evaluation of a TF-IDF and logistic regression classifier's robustness to adversarially obfuscated phishing emails at test time.

[View results]

OCR Text Classification: DistilBERT vs Longformer-DeBERTa

2026 · Model Benchmarking

A comparison of transformer architectures on OCR-extracted document classification by F1 score, latency, and memory footprint.

[View results]

Publications & Presentations

Ahmed, S. et al. (2023). "Synthesis and characterization study." Inorganic Chemistry Communications.

Published

Ahmed, S. et al. (2022). "Modeling and simulation of solar cells." Optics & Laser Technology.

Published

Ahmed, S. "Federated learning + synthetic data for vaccination equity research."

Manuscript in preparation

"DHS Vaccination Coverage Analysis." 2024 MSU Student Research Symposium: analysis of Demographic and Health Surveys data to identify vaccination coverage gaps across countries.

Conference presentation

Selected Projects

  • OpenDataBD: An open data initiative building research infrastructure for public government data on Bangladesh.[Visit]
  • Case Studies: Applied work in machine learning, healthcare analytics, and data infrastructure.[View]