2026-10-06T07:50:43Z https://api.figshare.com/v2/oai

oai:figshare.com:article/34046532 2026-10-01T15:17:01Z category_28813 category_29164 item_type_1 month_year_10_2026

Sampling-strategy comparison for the semi-infinite-domain inverse problem Gregorio Perez Bernal (22258905) Oscar Rincón Cardeño (22258780) Silvana Montoya-Noguera (22258901) Nicolas Guarin-Zapata (6532391) Applications in physical sciences Neural networks Scientific machine learning (SciML) Physics-informed neural networks (PINNs) Kolmogorov-Arnold Networks(KAN) Inverse problems for differential equations Unbounded Domains (UDs) <p dir="ltr">Uniform and exponential sampling are evaluated through the predicted fields, absolute errors, and MAE values measured inside, outside, and across the complete evaluation domain. The training region is [−5, 5] × [−5, 0] within the evaluation domain [−10, 10] × [−10, 0], with the top-boundary samples highlighted in blue.</p> 2026-10-01T15:17:01Z Image Figure 10.6084/m9.figshare.34046532.v1 https://figshare.com/articles/figure/Sampling-strategy_comparison_for_the_semi-infinite-domain_inverse_problem/34046532 CC BY 4.0