Develops and analyzes hypocoercivity-preserving C0-IP finite-element methods in space with hp-DG time stepping for kinetic Fokker-Planck equations on R^d x R^d, proving exponential decay via weighted Poincare inequalities and new trace estimates.
Acta Numer
4 Pith papers cite this work. Polarity classification is still indexing.
years
2026 4verdicts
UNVERDICTED 4representative citing papers
Infinite order-2 digital sequences over F_2 attain the optimal periodic L2-discrepancy bound of order C_d (log N)^{d/2}/N for all N except 1, improving prior order-5 constructions by reducing dimension from 5d to 2d.
Adaptive MFML algorithm saturates accuracy at low fidelities before escalating, cutting data costs up to 30x vs single-fidelity and 5x vs standard MFML on coupled cluster and excitation energies.
CUTS-GPR performs numerically exact Gaussian process regression with near-linear scaling in training points N and low-order polynomial scaling in dimensions D by exploiting additive kernels on incomplete grids.
citing papers explorer
-
Improvise, Adapt, Overcome: An On-The-Fly Multifidelity Algorithm for Efficient Machine Learning
Adaptive MFML algorithm saturates accuracy at low fidelities before escalating, cutting data costs up to 30x vs single-fidelity and 5x vs standard MFML on coupled cluster and excitation energies.
-
Don't Get Your Kroneckers in a Twist: Gaussian Processes on High-Dimensional Incomplete Grids
CUTS-GPR performs numerically exact Gaussian process regression with near-linear scaling in training points N and low-order polynomial scaling in dimensions D by exploiting additive kernels on incomplete grids.