We developed a novel uncertainty-aware framework MatPhase to predict material phases of electrodes from low contrast SEM images.
Simulations of red blood cells are important for a variety of biomedical applications, ranging from studies of blood diseases to the transport of circulating tumor cells.
A group of ORNL researchers and collaborators have been working to develop a pipeline that simulates radiotherapy across different scales, e.g., the individual cellular scale, multicellular/tissue scale, organ scale, and whole-body scale.
Members and students of the Computational Urban Sciences group demonstrated a method for generating scenarios of urban neighborhood growth based on existing physical structures and placement of buildings in neighborhoods.
A web-based GUI for INTERSECT has been created which allows a user to configure an experiment on an electron microscope, setting such parameters as maximum number of steps for the machine learning algorithm to perform.
A graph convolutional neural network (GCNN) was trained to accurately predict formation energy and mechanical properties of solid solution alloys crystallized in different lattice structures, thereby advancing the design of alloys for improving mechanic
In this work we focus on dynamics problems described by waves, i.e. by hyperbolic partial differential equations.
A research team from ORNL, Pacific Northwest National Laboratory, and Arizona State University has developed a novel method to detect out-of-distribution (OOD) samples in continual learning without forgetting the learned knowledge of preceding tasks.
ORNL researchers developed a novel nonlinear level set learning method to reduce dimensionality in high-dimensional function approximation.
The team conducted numerical studies to demonstrate the connection between the parameters of neural networks and the stochastic stability of DMMs.