Mlinaric, PetarGugercin, Serkan2023-12-212023-12-212023-04-261064-8275https://hdl.handle.net/10919/117252We provide a unifying framework for L-optimal reduced-order modeling for linear time-invariant dynamical systems and stationary parametric problems. Using parameter-separable forms of the reduced-model quantities, we derive the gradients of the L cost function with respect to the reduced matrices, which then allows a nonintrusive, data-driven, gradient-based descent algorithm to construct the optimal approximant using only output samples. By choosing an appropriate measure, the framework covers both continuous (Lebesgue) and discrete cost functions. We show the efficacy of the proposed algorithm via various numerical examples. Furthermore, we analyze under what conditions the data-driven approximant can be obtained via projection.Pages A554-A57825 page(s)application/pdfenIn Copyrightreduced-order modelingparametric stationary problemslinear time-invariant systemsoptimizationG2 normnonlinear least squaresArticle - Refereedhttps://doi.org/10.1137/22M1500678452Mlinaric, Petar [0000-0002-9437-7698]1095-7197