My research focuses on deterministic global optimization, with particular interest in the theory and algorithms of relaxation construction for large-scale nonconvex factorable problems. I work on enclosure arithmetics, McCormick relaxations, superposition relaxations, exploitation of separable structure, and reduced-space formulations for global optimization. Broadly speaking, my goal is to develop optimization methods that better balance tightness, expressiveness, and scalability, especially for nonconvex problems over wide, high-dimensional, and deeply composed domains.

More recently, my work has centered on superposition relaxation arithmetic and its augmented variants, with an emphasis on how structural properties such as convexity, supermodularity, and max-plus algebra induce exploitable forms of additivity and subadditivity. Through this perspective, I aim to build tighter and more compact reduced-space relaxations and, more broadly, to develop a systematic methodological framework for exploiting separability in deterministic global optimization.