arXiv subjects

Pedro Viegas

Publications and source records attributed to Pedro Viegas.

3 recordsLinked to original sources

arXiv2026

Stochastic hierarchical data-driven optimization: application to plasma-surface kinetics

This work introduces a stochastic hierarchical optimization framework inspired by Sloppy Model theory for the efficient calibration of physical models. Central to this method is the use of a reduced Hessian approximation, which identifies and targets the stiff parameter subspace using minimal simulation queries. This strategy enables efficient navigation of highly anisotropic landscapes, avoiding the computational burden of exhaustive sampling. To ensure rigorous inference, we integrate this approach with a probabilistic formulation that derives a principled objective loss function directly from observed data. We validate the framework by applying it to the problem of plasma-surface interactions, where accurate modelling is strictly limited by uncertainties in surface reactivity parameters and the computational cost of kinetic simulations. Comparative analysis demonstrates that our method consistently outperforms baseline optimization techniques in sample efficiency. This approach offers a general and scalable tool for optimizing models of complex reaction systems, ranging from plasma chemistry to biochemical networks.

cs.LG↗

arXiv2026

Conditional Denoising Model as a Physical Surrogate Model

Surrogate modeling for complex physical systems typically faces a trade-off between data-fitting accuracy and physical consistency. Physics-consistent approaches typically treat physical laws as soft constraints within the loss function, a strategy that frequently fails to guarantee strict adherence to the governing equations, or rely on post-processing corrections that do not intrinsically learn the underlying solution geometry. To address these limitations, we introduce the {Conditional Denoising Model (CDM)}, a generative model designed to learn the geometry of the physical manifold itself. By training the network to restore clean states from noisy ones, the model learns a vector field that points continuously towards the valid solution subspace. We introduce a time-independent formulation that transforms inference into a deterministic fixed-point iteration, effectively projecting noisy approximations onto the equilibrium manifold. Validated on a low-temperature plasma physics and chemistry benchmark, the CDM achieves higher parameter and data efficiency than physics-consistent baselines. Crucially, we demonstrate that the denoising objective acts as a powerful implicit regularizer: despite never seeing the governing equations during training, the model adheres to physical constraints more strictly than baselines trained with explicit physics losses.

cs.LG↗

arXiv2022

Benchmarking between fluid and global models for low-pressure oxygen DC glow discharges

This work focuses on the benchmarking between a zero-dimensional (0D) global model (LoKI) and a one-dimensional (1D) radial fluid model for the positive column of oxygen DC glow discharges in a tube of 1 cm inner radius at pressures between 0.5 Torr and 10 Torr. The data used in the two models are the same, so that the difference between the models is reduced to dimensionality. A good agreement is found between the two models on the main discharge parameters, with relative differences below 5%. The agreement on species average number densities, charged and neutral, is slightly worse, with relative differences increasing with pressure from 11% at 0.5 Torr to 57% at 10 Torr. The success of the 0D global model in describing these plasmas through volume averaged quantities decreases with pressure, due to pressure-driven narrowing of radial profiles. Hence, in the studied conditions, we recommend the use of volume-averaged models only in the pressure range up to 10 Torr.

physics.plasm-ph↗