For computer models of complex physical and engineering systems, surrogates are often necessary to accelerate analyses that require large numbers of model evaluations. When a computer model is stochastic, its surrogate should also be stochastic. …
Subspace-constrained Mean Shift (SCMS) is an iterative algorithm for non-parametric ridge estimation. Although SCMS is computationally feasible, it is only linearly convergent and its computational complexity per iteration is quadratic in sample size …
When data points exhibit salient geometric structure, density estimation and generative modeling can be more efficient by exploiting the data manifold. Here we propose Manifold Scaffold, a generative model for data concentrated near a manifold. First …
For probability density estimation on Riemannian manifolds, many methods have been proposed, parametric or non-parametric. However, whether they can be implemented for general manifolds is in question. Here we propose a kernel-based method for …