Abstract Neuro-symbolic AI presents tremendous opportunities for more efficient foundation models in the form of integrating classical symbolic methods with deep learning. However, that integration is faced with several key challenges, including the correspondence between symbolic and neural systems and the scale of modern AI models. In this talk, we cover some of these challenges […]
Abstract: Bulk nanostructured materials offer tremendous opportunity for re-inventing materials, but also pose challenges both in terms of characterization, design, processing, and scaling. This talk will present recent work developing nanoengineered hierarchical advanced (aerospace-grade) composites with enhanced mechanical properties, with a focus on imparting multifunctionality. Such hybrid advanced composites employ aligned nanofibers (in most of […]
Join Duke's Pratt School of Engineering during this week-long event celebrating how engineers make a difference in our world. More at pratt.duke.edu/e-week
Artificial intelligence and machine learning have become important components of the computational toolbox that can be used to advance chemical research and discovery. Our group has been working on advancing AI/ML as it applies to the broad subfields of synthetic organic chemistry (for developing/discovering reactions), medicinal chemistry (for developing/discovering new functional molecules), and analytical chemistry […]
Abstract: During cardiovascular development, peristaltic contraction of the embryonic heart tube produces time-varying hemodynamic forces and pressure gradients across the atrioventricular canal. However, the relative importance of myocardial contraction and hemodynamic force to modulate cardiovascular morphogenesis in the non-Newtonian flow regime remains poorly understood. By developing the 4-D light-sheet fluorescent microscope and post-imaging machine-learning algorithms, […]
