ML for Scientific Data - Research Highlights

Machine Learning for Scientific Data Analytics at DOE User Facilities

Sponsor: DOE - Office of Advanced Scientific Computing Research

Highlights on foundational SciML methods

PI3NN: Out-of-distribution-aware prediction intervals from three neural networks

A stochastic backpropagation for quantifying uncertainty in deep learning

Kernel learning-based backward SDE filter for data assimilation

Exploiting the Parabolic Loss Landscape to Accelerate Black-Box Adversarial Attack

Learning Green’s functions of linear reaction-diffusion equations

A one-stage domain adaptation network for semantic segmentation

Style mixing and patchwise prototypical matching for adaptive semantic segmentation

Discovering State Variables Hidden in Experimental Data

Learning nonlinear level sets for dimensionality reduction in function approximation

Highlights on Scientific Impacts

Model Calibration of the Liquid Mercury Spallation Target at SNS

Machine-Learning-Assisted Automation of Single Crystal Neutron Diffraction

Atomic Resolution Cryogenic 4D-STEM Imaging via Robust Distortion Correction

A scalable transformer model for real-time decision making in neutron scattering experiments

Rapid optimization of crystal field Hamiltonian

Page updated

Google Sites

Report abuse