This paper extends MST-Direct for multivariate geostatistical simulation from bivariate, unconditional, small-grid to multivariate, conditional, and large-grid settings. It tackles scalability, multiple variables, and hard-data conditioning while preserving the multivariate joint distribution.
This paper investigates Bayesian Optimization for expensive, black-box, non-convex optimization problems with symmetries. It proposes an Optimal Transport-based Permutation-Invariant Bayesian Optimization approach, specifically applied to the layout design of offshore wind farms.
This paper introduces FastSinkhorn, a native CUDA implementation of the log-domain Sinkhorn algorithm that provides faster and more stable solutions for optimal transport (OT) problems. It achieves a 12x speedup over the POT library and 5.9x over GPU-accelerated PyTorch baselines, maintaining numerical stability for small regularization parameters.