|
Combinatorics And Artificial Intelligence LaboratoryCAAIL |
About | Research Themes | Software | Funding | Members | Projects | Publications | Opportunities
About
The Combinatorics And Artificial Intelligence Laboratory (CAAIL) at the Lebanese American University conducts research at the intersection of combinatorics, graph algorithms, artificial intelligence, and data science. Our work spans exact, parameterized, and enumeration algorithms, graph modification, clustering and biclustering, feature selection, graph compression, network analysis, and explainable AI. While rooted in the design and analysis of efficient algorithms, the laboratory increasingly applies these methods to problems arising in biomedicine, scientific discovery, cybersecurity, and other data-rich domains where interpretable and computationally efficient approaches can provide practical value.
Research Themes
| Lightweight and Explainable AI | Efficient AI methods that use combinatorial structure, compact representations, and transparent decision rules instead of relying only on large black-box models. |
| Biomarker Discovery and Feature Selection | Graph-based and combinatorial methods for identifying small, interpretable sets of predictive features in biomedical and high-dimensional data, including applications to cancer and disease prediction. |
| Clustering, Biclustering, and Overlapping Structures | Algorithms and heuristics for discovering clusters, biclusters, and overlapping communities in complex data, including graph modification and vertex-splitting approaches. |
| Domination, Covering, and Packing in Networks | Structural and algorithmic studies of domination, covering, packing, and representative-set problems, with applications to networks, feature selection, and influence analysis. |
| Graph Compression and Representation | Compact representations of large graphs and networks, including graph compression, embedding, dimensionality reduction, and multi-intersection graph representations. |
| Exact, Parameterized, and Enumeration Algorithms | Design and analysis of exact, parameterized, kernelization, and enumeration algorithms for hard graph and network problems, with emphasis on both theoretical guarantees and practical performance. |
| Data Security, Privacy, and Trustworthy Computing | Algorithmic approaches to privacy, anonymization, secure data analysis, and trustworthy AI, especially in networked and data-sensitive environments. |
Software and Code Portal
|
CAAIL Code Portal -- Coming Soon The laboratory is preparing a software portal that will provide public access to selected research implementations. The portal will initially feature a Vertex Cover Solver and will gradually expand to include additional tools for graph algorithms, optimization, clustering, and feature selection. Users will be able to upload modest-size graph instances, run algorithms under predefined resource limits, and obtain downloadable results. The portal is intended for demonstration, reproducibility, teaching, and small experimental studies. First planned release: Vertex Cover Solver. Repository: GitHub link to be added Portal: Upload/testing page coming soon |
Funding and Support
Research conducted by CAAIL has been supported through:
- Three successive France-Lebanon PHC CEDRE grants.
- Two LAU President's Intramural Research Fund (PIRF) awards.
- Participation in ORNL/DOE-funded collaborative research projects.
Lab Members
Director
Research Associates
Graduate and Undergraduate Researchers
|
|
Selected Projects and Current Activities
The project pages below include related publications and additional project information.
- Algorithmic and Applied Aspects of Network Domination
- Combinatorial Methods for Feature Selection and Biomarker Discovery
- Machine Learning for Optimization Problems
- Correlation Clustering via Graph Modification
- Graph Compression and Embedding
Current Research Projects
| DNA fragment assembly | Reina Najjar |
| Protein data analysis | Lucas Isenmann, Mira Madi and Aya Darwish |
| Feature Selection and Biomarker Discovery for Disease Prediction | Cynthia Chaaya and Nadim Obeid |
| Machine learning for optimization problems | Omar Kaddoura and Joseph Jabbour |
| Clustering and biclustering with overlaps | Lucas Isenmann, Sergio Thoumi and Zeina Mershad |
| Vertex splitting into special graph classes | Lucas Isenmann, Sergio Thoumi, Mira Madi, Jean-Paul Khawam and Joe Semaan |
| Exact, parameterized, and enumeration algorithms | Lucas Isenmann, Sergio Thoumi, Reina Najjar and Joseph Jabbour |
| Stochastic thresholding for graph-based clustering | Mohammad El-Asal |
| Network data privacy and security | Abdallah Srayeldine |
| LLMs and blockchain systems | Mohamad Sabra |
| Blockchain consensus algorithms | Jad El Masri |
Recent Highlights
- Novel feature-selection methods for compact biomarker discovery and high-dimensional biomedical data.
- New exact, parameterized, and enumeration algorithms for graph and network problems.
- Vertex-splitting frameworks for structure discovery, graph modification, clustering, and biclustering.
- Research on domination, covering, and packing methods for network analysis and optimization.
- Public software portal currently under development for modest-size algorithmic experiments.
Opportunities
|
We welcome collaborations and student participation. Students interested in graph algorithms, AI, biomedical data analysis, clustering, feature selection, or software implementation are encouraged to contact the lab director. We also welcome inquiries from potential postdoctoral researchers and external collaborators. Email: faisal.abukhzam(AT)lau.edu.lb |
Last updated: 2026-06-15