Subject to: Bissan Ghaddar

2025-04-17 · Source: Subject to · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Mathematics & Computational Sciences, Operations & Process Management · Depth: Advanced, extended

Summary

Bissan Ghaddar, the John M. Thomson Chair in Engineering Leadership and Innovation and an associate professor at the Ivey Business School, discusses her career journey and research at the intersection of machine learning and nonlinear optimization. Born in Beirut, Lebanon, in 1983, Ghaddar grew up in a resilient environment shaped by the Lebanese Civil War. She pursued computer engineering at the American University of Beirut, where she was introduced to operations research. Her Master's work focused on semi-definite programming (SDP) for the minimum K partition problem, which she presented at ICC Opt in 2006. After her PhD in polynomial optimization, she worked at the Department of National Defense in Canada and IBM Research in Dublin, gaining industrial experience in applying optimization to real-world problems like water and energy networks. Now at Ivey Business School, her current research focuses on optimizing smart sustainable cities, including energy systems, water networks, transportation, and telecom, often integrating machine learning with optimization techniques to solve large-scale problems.

Key takeaway

For AI Scientists and Research Scientists working on complex optimization problems, consider integrating machine learning with traditional optimization methods like SDP and polynomial optimization. Focus on developing methodologically sound approaches that offer interpretability and guarantees, rather than relying on black-box solutions, to ensure practical applicability and robust contributions to the field. Your ability to bridge theoretical rigor with real-world impact will be key.

Key insights

Resilience and adaptability are crucial for navigating complex academic and professional paths, especially in optimization.

Principles

• Exploit problem structure for computational efficiency.
• Combine theoretical methods with practical applications.
• Interdisciplinary collaboration enhances problem-solving.

Method

Integrate machine learning with nonlinear optimization by developing tight, cheap relaxations and using learning techniques for branching rules and cut generation in spatial branch-and-bound algorithms.

In practice

• Apply SDP relaxations to sparse network problems.
• Use machine learning to optimize branching rules.
• Frame research for diverse technical and business audiences.

Topics

• Nonlinear Optimization
• Semi-Definite Programming
• Polynomial Optimization
• Smart Sustainable Cities
• Machine Learning Integration

Best for: AI Scientist, Research Scientist, AI Student

Related on AIssential

• Subject to: Jorge Nocedal — Nocedal's journey highlights the impact of diverse experiences and persistent exploration in shaping foundational contributions to optimization.
• Subject to: Laureano Escudero — Laureano Escudero's career exemplifies dedication to mathematical optimization and its real-world applications.
• On the Oracle Complexity of Interpolation-Based Gradient Descent — PPI-GD improves gradient descent oracle complexity by approximating gradients via polynomial interpolation over data domain patches.
• IEEE Transactions on Evolutionary Computation, Volume 30, Issue 1, February 2025 — Evolutionary computation continues to advance, integrating with AI and deep learning for complex optimization and scheduling tasks.
• Learning with Simulators: No Regret in a Computationally Bounded World — Simulators enable classical learning guarantees for complex, dependent data, broadening the PAC model.
• Constrained Hybrid Metaheuristic: A Universal Framework for Continuous Optimisation — cHM is a universal, adaptive hybrid metaheuristic that dynamically combines diverse optimizers for robust continuous optimization.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Subject to.