Special Session 8:
Graph Machine Learning for Complex Systems and Applications
Recent advances in machine learning increasingly require models that can reason over complex systems characterized by rich relational structure, interdependencies, and higher-order interactions. Many real-world domains, including biological systems, healthcare processes, social and information networks, engineered infrastructures, and multimodal environments, cannot be adequately represented using traditional vector- based learning paradigms. Graph-based representations and learning frameworks have therefore emerged as a foundational methodology for capturing structural, relational, and contextual information that is essential for robust and interpretable decision making.
This special session aims to provide a focused forum for researchers and practitioners working on graph machine learning methods and their applications to complex systems. The session will cover methodological advances in graph neural networks, including higher-order, hypergraph, and heterogeneous graph learning, as well as structured relational modeling. It will also address practical challenges related to oversmoothing, oversquashing, scalability, robustness, and generalization in real-world settings. A particular emphasis will be placed on explainable and trustworthy graph-based learning, recognizing the growing need for transparent and interpretable models in high-stakes application domains. In addition, the session will highlight emerging research directions that integrate graph learning with large language models (LLMs) and knowledge graph reasoning. Such hybrid approaches enable more expressive reasoning over structured and multimodal data, bridging symbolic knowledge representations with data-driven learning. By combining graph-based inductive biases with the representational power of foundation models, these approaches offer promising pathways toward more context-aware, explainable, and reliable machine learning systems.
Overall, the session emphasizes applied machine learning and encourages contributions that demonstrate how graph-based models and graph-enhanced LLM frameworks can improve predictive performance, interpretability, and decision support across diverse complex domains. By connecting theoretical developments with practical applications, the session aligns closely with ICMLA’s mission to advance both foundational research and impactful real-world deployment of machine learning technologies.
Scope and topics:
Topics of interest include, but are not limited to:
- Graph Neural Networks and Graph Representation Learning
- Oversmoothing and Oversquashing
- Machine Learning on Complex, Heterogeneous Relational and Structured Data
- Graph-Based Models for Biological Healthcare and Networked Systems
- Explainable, Trustworthy, and Robust Graph Machine Learning
- Higher Order Hypergraph and Set-Based Learning Models
- Scalability, Generalization, and Robustness in Graph Learning
- Integration of LLMs with Graph and Knowledge Graph Representations
- Knowledge Graph Reasoning Completion and Inference with Machine Learning
- LLM Enhanced Graph Learning for Structured and Multimodal Data
- Multimodal Learning with Graph Representations
- Applications of Graph Machine Learning in Engineering, Security, and Decision Support
Chairs:
- Chair Emails
- Biographies
Chair: Khaled Mohammed Saifuddin: k.saifuddin@siu.edu
Co-Chair: Abdur Rahman Bin Shahid:abdurrahmanbin.shahid@siu.edu
Co-Chair: Alvi Ataur Khalil:a.khalil@siu.edu
Dr. Khaled Mohammed Saifuddin is a tenure-track Assistant Professor in the Department of Computer Science at SIUC. His research focuses on graph machine learning, network science, and AI-driven modeling of biological and complex systems, emphasizing higher-order relational learning and interpretability. He serves as a reviewer and program committee member for major conferences and journals, including ACM SIGKDD (KDD), IJCNN, WWW, PAKDD, ICMLA, and Nature Communications Biology. He serves as Workshop Area Chair for the NeurIPS and ICML Workshops on AI4Science and as Proceedings Chair of the SNAS Interdisciplinary Research Conference.
Dr. Abdur Rahman Bin Shahid is an Assistant Professor at SIUC. His research focuses on cybersecurity, deep learning, adversarial ML, multimodal AI, usable security and privacy, generative AI, Cyber-Physical Systems, and Internet of Things. Dr. Shahid served as a co-chair for special sessions at IEEE ICMLA 2025 and for the International Workshop on Security, Privacy, and Trust for Emergency Events (Emergen- cyComm), held in conjunction with SecureComm 2020. He also served as a program committee member of several conferences, including IEEE CCNC, IEEE FIE, IEEE SSCI, and a reviewer for numerous journals, including IEEE IoT Journals, IEEE TII, and IEEE TDSC.
Dr. Alvi Ataur Khalil is a tenure-track Assistant Professor in the Department of Computer Science, School of Computing, at SIUC. Dr. Khalil’s research focuses on blockchain security, particularly off-chain Layer-2 vulnerabilities and defenses, intelligent UAV control using reinforcement learning, and AI-driven cybersecurity solutions for cyber-physical systems. He has authored over twenty peer-reviewed papers published in leading venues, including IEEE TNSM, Elsevier Computer Networks, IEEE/IFIP DSN, ACSAC, EAI SecureComm, IEEE CNS, ACM DLT, IEEE LCN, CNSM, IEEE SMARTCOMP, and IEEE COMPSAC.
Paper Submission Instructions
All papers will be double-blind reviewed and must present original work.
- CMT Submission Site
- Select the track: Special Session 8: Graph Machine Learning for Complex Systems and Applications
Papers submitted for reviewing should conform to IEEE specifications. Manuscript templates can be downloaded from:
Keydates
- Submission Deadline: June 20, 2026
- Notification of Acceptance: July 10, 2026
Registration
In order for your paper to be published in the proceedings you must register to the conference.
Paper Presentation Instructions
The papers submitted to this track will be presented in person as part of the conference. There is no virtual presentation for this session.
ICMLA'26