Machine Learning for Complex Networks: Generalization, Fairness, and Model Selection in Graph-Based Systems Across Social and Biological Domains

Aug 5, 2026

Networks provide a foundational framework for modeling complex systems across biological interactions, social structures, economic markets, and information networks. Machine learning improves pattern extraction and predictive performance in graph-based applications such as market design, recommendation, and biomedical discovery. Key challenges remain, especially overreliance on simplified models and synthetic benchmarks that ignore structured dependencies and temporal, structured missingness. This dissertation develops fairness-aware approaches for stable matching with correlated preferences, a principled model selection method for signed networks using GRASMOS, and improved link-prediction evaluation with temporal evolution to reduce generalization gaps.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages132

Reading Time5 h 33 min

AccessFree to download

Abstract

Fairness in stable matching with correlated preferences

Model selection for signed networks and GRASMOS

Temporal evaluation for link prediction

Rochester Institute of Technology RIT Digital Institutional Repository Theses 3-2025 Machine Learning for Complex Networks: Generalization, Fairness, and Model Selection in Graph-Based Systems Across Social and Biological Domains Angelina Brilliantova[lb9849@rit.edu](lb9849@rit.edu) Follow this and additional works at: [https://repository.rit.edu/theses](https://repository.rit.edu/theses) Recommended Citation Brilliantova, Angelina, "Machine Learning for Complex Networks: Generalization, Fairness, and Model Selection in Graph-Based Systems Across Social and Biological Domains" (2025) . Thesis. Rochester Institute of Technology. Accessed from This Dissertation is brought to you for free and open access by the RIT Libraries. For more information, please contact [repository@rit.edu](repository@rit.edu). Machine Learning for Complex Networks: Generalization, Fairness, and Model Selection in Graph-Based Systems Across Social and Biological Domains by Angelina Brilliantova A dissertation submitted in partial fulﬁllment of the requirements for the degree of Doctor of Philosophy in Computing and Information Sciences B. Thomas Golisano College of Computing and Information Sciences Rochester Institute of Technology Rochester, New York March, 2025 Machine Learning for Complex Networks: Generalization, Fairness, and Model Selection in Graph-Based Systems Across Social and Biological Domains by Angelina Brilliantova Committee Approval: We, the undersigned committee members, certify that we have advised and/or supervised the candidate on the work described in this dissertation. We further certify that we have reviewed the dissertation manuscript and approve it in partial fulﬁllment of the requirements of the degree of Doctor of Philosophy in Computing and Information Sciences. | Dr. Ivona Bez´akov´a
Dissertation Advisor | Date | | --- | --- | | Dr. Stanisław Radziszowski
Dissertation Committee Member | Date | | Dr. Varsha Dani
Dissertation Committee Member | Date | | Dr. Rui Li
Dissertation Committee Member | Date | | Dr. Anurag Agarwal
Dissertation Defense Chairperson
Certiﬁed by: | Date | Dr. Pengcheng Shi Date Ph.D. Program Director, Computing and Information Sciences iii ©2025 Angelina Brilliantova All rights reserved. Machine Learning for Complex Networks: Generalization, Fairness, and Model Selection in Graph-Based Systems Across Social and Biological Domains by Angelina Brilliantova Submitted to the B. Thomas Golisano College of Computing and Information Sciences Ph.D. Program in Computing and Information Sciences in partial fulﬁllment of the requirements for the Doctor of Philosophy Degree at the Rochester Institute of Technology Abstract Networks serve as a fundamental framework for modeling complex systems across diverse domains, from biological interactions and social structures to economic markets and information networks. Machine learning has signiﬁcantly enhanced the ability to extract meaningful patterns from graph-based systems, improving predictive accuracy in applications such as market design, recommendation systems, and biomedical discovery. Despite the successes of machine learning in analyzing complex networks, several fundamental challenges remain. A major issue is the reliance on simplistic models for theoretical analysis, simulation, and benchmarking, which often fail to capture the complexities of real-world networked systems. Many theoretical studies focus on oversimpliﬁed generative models that assume uniform randomness in edge formation, preference rankings, or network evolution, limiting their applicability to real-world problems. While these models provide analytical tractability, they often overlook the structured dependencies and correlations that shape real-world interactions. Similarly, the simulation and benchmarking of machine learning models for networks frequently rely on synthetic datasets that do n

FAQs

What problem does the dissertation address in machine learning for complex networks?

It targets limitations of simplified theoretical models and synthetic benchmarks that fail to capture structured dependencies and temporal, structured missingness in real-world networks.

How does the work handle fairness in stable matching?

It studies fairness under correlated preferences, analyzes how preference asymmetries affect fairness outcomes, and proposes efficient fairness-aware solutions that preserve stability.

What contributions are made to model selection and evaluation for graph tasks?

It introduces GRASMOS for maximum-likelihood model selection in signed networks to infer realistic sign assignments, and it improves link-prediction evaluation by incorporating temporal graph evolution to reduce generalization gaps.

Aug 5, 2026

Upload by River Wang