AI • Cybersecurity • Graph Machine Learning

Dr. Farahnaz Hosseini

Teaching Assistant Professor
School of Cybersecurity, Old Dominion University

I develop trustworthy and resilient artificial intelligence systems with a focus on AI security, adversarial machine learning, graph neural networks, and large-scale data analytics. My research connects advanced AI methods with cybersecurity, public health, socioeconomic modeling, and other high-impact applications.

Explore My Research Download CV

About Me

Trustworthy AI, graph intelligence, and cybersecurity for real-world impact.

I am a computer scientist and cybersecurity faculty member with a Ph.D. in Computer Science from the University of North Texas. My research spans trustworthy artificial intelligence, AI security, adversarial machine learning, graph neural networks, representation learning, predictive modeling, large-scale graph analytics, and high-performance computing.

I am particularly interested in developing AI systems that remain robust, secure, interpretable, and reliable under realistic and adversarial conditions.

My interdisciplinary research has included cybersecurity, socioeconomic prediction, public-health modeling, biomedical AI, dynamic networks, and responsible AI.

Contact Me

22+

Peer-Reviewed Publications

62+

Scholarly Citations

5

h-index

ODU

Teaching Assistant Professor

Current Academic Position

Teaching Assistant Professor

School of Cybersecurity
Old Dominion University
Norfolk, Virginia
August 2026 – Present

Teaching

Teach undergraduate and graduate cybersecurity courses spanning cyber defense, malware, network reconnaissance, security tools, the social and ethical dimensions of cybersecurity, and emerging AI-security topics.

Research

Develop research at the intersection of trustworthy AI, adversarial robustness, graph machine learning, cybersecurity, and high-impact data-driven applications.

Mentoring

Mentor students in AI-enabled cybersecurity, phishing detection, machine learning, natural language processing, cloud security, and data-driven cyber defense.

University Service

Contribute to university and School of Cybersecurity service, including academic assessment, hands-on course review, curriculum quality, and interdisciplinary collaboration.

Research Areas

Secure, trustworthy, and graph-based artificial intelligence.

Trustworthy AI & AI Security

Studying the security, reliability, robustness, and responsible deployment of AI systems in adversarial and high-impact environments.

Adversarial Machine Learning

Evaluating feature and structural attacks, model resilience, attack mitigation, and defense strategies for machine-learning and graph-learning systems.

Graph Neural Networks

Designing GCN, GraphSAGE, graph-augmented, and representation-learning approaches for structured data, prediction, and community discovery.

AI for Cybersecurity

Applying artificial intelligence and machine learning to phishing detection, cyber defense, anomalous behavior analysis, secure systems, and emerging cyber threats.

Large-Scale Graph Analytics & HPC

Developing scalable graph algorithms, including k-core and k-truss methods, for large dynamic networks using high-performance and parallel computing.

AI for Health & Society

Applying graph machine learning and predictive modeling to public health, socioeconomic data, biomedical applications, and geospatial decision support.

Featured Research Projects

Selected work across trustworthy AI, cybersecurity, graph learning, public health, and scalable computing.

Adversarial Robustness of Graph AI

Evaluating GCN, GraphSAGE, and hybrid graph-machine-learning models under feature, structural, and mixed adversarial perturbations.

Similarity-Consistency Filter (SCF)

A graph-defense strategy designed to identify and suppress suspicious graph connections and improve model resilience under adversarial structural attacks.

Graph AI for Public Health

Predictive modeling of diabetes, obesity, and mental-health indicators using socioeconomic, geographic, and graph-based data.

Socioeconomic Prediction with GNNs

Graph neural network and hybrid machine-learning approaches for predicting socioeconomic outcomes using large-scale U.S. Census and geospatial data.

Census2Vec

Geo-embedded representation learning for socioeconomic predictive modeling, integrating geographic structure with learned vector representations.

Dynamic Graph Analytics

Scalable k-core and k-truss algorithms for large dynamic networks using high-performance computing and parallel processing.

AI-Enabled Cybersecurity Research

Research and student mentoring in phishing detection, machine learning, natural language processing, cloud security, and data-driven cyber defense.

Biomedical AI

Research involving white-blood-cell detection, forensic age estimation, medical-image processing, and MRI-based brain tumor analysis.

Responsible & Emerging AI

Research on bias analysis in language models, prompt-based evaluation, responsible artificial intelligence, multimodal AI, and emerging generative AI systems.

Publications & Research Impact

22+ peer-reviewed publications and 62+ scholarly citations.

My publication record spans artificial intelligence, graph learning, cybersecurity, high-performance computing, biomedical AI, and data-driven predictive modeling.

Graph Neural Networks and Geo-Embedded Census Data for Socioeconomic Prediction: A Comprehensive Review

Farahnaz Hosseini and Ting Xiao, 2026.

Robust Socioeconomic Community Discovery from Census-Derived Indicators via KNN Graphs and Louvain Stability-Conductance Analysis

Farahnaz Hosseini and Ting Xiao, IEEE SoutheastCon 2026.

Discovering Socioeconomic Structure with Unsupervised Graph Neural Embeddings on Census Data

Farahnaz Hosseini and Ting Xiao, CSCI 2025.

Graph-Augmented Machine Learning for Socioeconomic Prediction Using Census Data

Farahnaz Hosseini and Ting Xiao, CSCI 2025.

Census2Vec: Enhancing Socioeconomic Predictive Models with Geo-Embedded Data

R. V. K. Bevara, I. Wagenvoord, F. Hosseini, et al., Springer Nature, 2024.

View complete publication record on Google Scholar →

Teaching & Mentoring

Cybersecurity education, AI instruction, and student research development.

Old Dominion University

Teach undergraduate and graduate cybersecurity courses, including Cybersecurity and the Social Sciences, Advanced Techniques for Cyber Defense, and Advanced Cybersecurity Techniques and Operations.

University of North Texas

Supported university instruction and project-based student work in computing, artificial intelligence, data analysis, and visualization.

Research Mentoring

Mentor students in AI-enabled cybersecurity, machine learning, NLP, phishing detection, graph algorithms, technical writing, and research planning.

Professional & University Service

Supporting academic quality, student learning, and research communities.

General Education Assessment Committee

Serve as a faculty member of Old Dominion University’s General Education Assessment Committee for the 2026–2027 academic year, contributing to university-wide assessment, academic quality, and continuous improvement initiatives.

Hands-on Review Committee

Serve on the School of Cybersecurity’s Hands-on Review Committee, contributing to the review of practical course activities, applied learning components, and hands-on cybersecurity instruction.

Professional Communities & Mentoring

Member of ACM, IEEE, Computing Research Association, AnitaB.org, and Women in High Performance Computing, with experience mentoring undergraduate and graduate researchers.

Contact

Let’s connect.

I welcome research collaborations, interdisciplinary AI projects, student mentoring opportunities, and conversations around trustworthy AI, AI security, graph machine learning, and cybersecurity.

fhossein@odu.edu LinkedIn Google Scholar Download Academic CV